Metrics (KPI's) to Assess Process Performance

Ab ela rd o A. Flores Flores Ja m es W. Mc Ca herty Revision #03 – March 27, 2007 Caterpillar Confidential Yellow Copyright 2007 Caterpillar. Caterpillar. All Rights Reserved. Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Introduction The p rima ry go a l a nd resp esp on sib ili ility of the M a intena nc e Dep a rtm en t is is to maintain equipment in optimum condition via problem / failure avo id anc e. One of the key key tools tools of the the Maintena Maintena nc e Dep Dep a rtment is is pe rforma forma nc e mea sureme ureme nt. In ord ord er to be effective, pe rforma forma nc e me a surem urem ents should g ive us not o nly nly an a c c ura ura te p ic ture ture o f how w e a re performing relative to established targets and/or global benchmarks but also also p rovide ovide us with a p red ic tion tion o r projec projec tion tion a s to w hat we c a n expec t in te rms of future future p erformanc erformanc e. In spite of the fact that it is human nature to want to report good news, the most valuable performance measures assist us in the identification of weak areas, poor practices and other opportunities for improvement. Therefore, herefore, itit b eho ove s us to ta ke a c ritic tic a l loo loo k at our pe rforma forma nc e via a n honest assessment of that performance. Only in this way are we able to c orrec orrec t sma sma ll issues b efo re the y be c om e ma jor p rob lem s. The c ollec ollec tion of performance measures included in this document help us to identify a nd und ersta ersta nd on-si on-site p erforma erforma nc e iss issues relate d to p rojec t he a lth a nd ma intena nc e / rep a ir p roc ess ess p erforma erforma nc e in in sup sup p ort of that p rojec t. In doing so we are also able to gain some insight into related issues such as application severity, operating practices, contract guarantees / c om mitments mitments,, and c ontrac ontrac t financ financ ial hea lth. “Performance Metric” is a term used to describe the outcome of any p roc ess ess used used to c ollec ollec t, ana lyze, yze, interpr interpret et a nd p resent esent q ua ntita ntita tive tive d a ta . It is a measurement parameter that enables performance against some p re-de fined fined Ta rg et o r Benc hma rk to b e m onitored onitored … a me a surem urem ent used to gauge performance of a function, operation or business relative to p a st res resul ults ts a nd p rojec ted future future b eha vior. vior. “Benchmark” is another term frequently used to describe performance. Benchmark is defined as a world-class performance standard relative to a specific performance metric. A Benchmark represents and quantifies "wo rld -c la ss p erforma erforma nc e or be st p ra c tic tic e" of a n o p eration or of sp sp ec ific fic functions within that operation according to a specified performance metric. A benchmark is determined by and represents actual, documented, sustainable performance over time relative to some pe rforma forma nc e me tri tric .

Caterpillar Confidential Yellow Copyright 2007 Caterpillar. Caterpillar. All Rights Reserved. Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Introduction The p rima ry go a l a nd resp esp on sib ili ility of the M a intena nc e Dep a rtm en t is is to maintain equipment in optimum condition via problem / failure avo id anc e. One of the key key tools tools of the the Maintena Maintena nc e Dep Dep a rtment is is pe rforma forma nc e mea sureme ureme nt. In ord ord er to be effective, pe rforma forma nc e me a surem urem ents should g ive us not o nly nly an a c c ura ura te p ic ture ture o f how w e a re performing relative to established targets and/or global benchmarks but also also p rovide ovide us with a p red ic tion tion o r projec projec tion tion a s to w hat we c a n expec t in te rms of future future p erformanc erformanc e. In spite of the fact that it is human nature to want to report good news, the most valuable performance measures assist us in the identification of weak areas, poor practices and other opportunities for improvement. Therefore, herefore, itit b eho ove s us to ta ke a c ritic tic a l loo loo k at our pe rforma forma nc e via a n honest assessment of that performance. Only in this way are we able to c orrec orrec t sma sma ll issues b efo re the y be c om e ma jor p rob lem s. The c ollec ollec tion of performance measures included in this document help us to identify a nd und ersta ersta nd on-si on-site p erforma erforma nc e iss issues relate d to p rojec t he a lth a nd ma intena nc e / rep a ir p roc ess ess p erforma erforma nc e in in sup sup p ort of that p rojec t. In doing so we are also able to gain some insight into related issues such as application severity, operating practices, contract guarantees / c om mitments mitments,, and c ontrac ontrac t financ financ ial hea lth. “Performance Metric” is a term used to describe the outcome of any p roc ess ess used used to c ollec ollec t, ana lyze, yze, interpr interpret et a nd p resent esent q ua ntita ntita tive tive d a ta . It is a measurement parameter that enables performance against some p re-de fined fined Ta rg et o r Benc hma rk to b e m onitored onitored … a me a surem urem ent used to gauge performance of a function, operation or business relative to p a st res resul ults ts a nd p rojec ted future future b eha vior. vior. “Benchmark” is another term frequently used to describe performance. Benchmark is defined as a world-class performance standard relative to a specific performance metric. A Benchmark represents and quantifies "wo rld -c la ss p erforma erforma nc e or be st p ra c tic tic e" of a n o p eration or of sp sp ec ific fic functions within that operation according to a specified performance metric. A benchmark is determined by and represents actual, documented, sustainable performance over time relative to some pe rforma forma nc e me tri tric .


Caterpillar has invested a great deal of time, energy and resources identifying and developing several metrics of performance that quantify and trend product and project health. Based on that experience we have been able to document actual site performance in many of those areas and feel very confident representing some of those results as Benchmarks. Unfortunately, the overwhelming majority of that effort has been c on c en tra tra ted on flee ts of large O ff Hi Hig hw a y Truc ks. ks. The refo re, mo st of the Benchmarks used in this document are specific to those fleets. When we have felt comfortable in doing so, we have also cited our best estimates of w ha t those those Benc Benc hma rks might b e for other mini mining ng eq uip uip me nt ifif we ha d access to a significant sample of data for those machines. In situations where we do not have that same level of confidence, we have made tha t c lea r a s we d o no t c a re to misl mislea d the user user b y sp sp ec ula ula ting ting . The m etric etric s c on ta ine d herein will will p rov id e insi insig ht into into t he p erforma erforma nc e o f the individual Maintenance & Repair Processes and their contributions to the en d results esults.. Tho se m e tric tric s a re o rg a nize nize d a s fo llow s:

1. - Fl Flee t Perfor Performa ma nc e Me tri tric s 2. - Ma inte na nc e & Rep a ir Pr Proc ess esses Metric Metric s 2.1 2.1 - Prev ent ive M a inte na nc e 2.2 2.2 - Co nd ition M on ito ring 2.3 2.3 - Ba c klog klog Ma na ge me nt 2.4 2.4 - Plann ing a nd Sc he d uling uling 2.5 2.5 - Pa rts Ma na g em ent 2.6 2.6 - Rep a ir Ma na g em ent 2.7 - Comp onent Mana gem ent 2.8 - Huma Hum a n Re Re so urc e s - Training ra ining 2.9 2.9 - Continuo us Imp rov em ent


1.1 1. 1 Flee Flee t Pe Pe rforma nc e Me tric tric s The fo llow ing m e a sures a re “ To p Tier” ier” p e rfo rm a nc e m e tric tric s o r Key Performance Indicators (KPI’s) that enable management to quantify, a ssess ess a nd mo nitor nitor produc t hea lth a nd site p erforma erforma nc e.

1.1.1 - M e a n Tim Tim e Be tw twe e e n Shut Shutd d o w ns (M TBS) Definition: The a vera vera ge op erating erating time time be twee n ma c hine hine stop stop pa ge s … a function function of the average frequency of equipment downtime events, expressed in hours. Ca lc ula ula tion tion Method ology:

M TBS =

Op erating erating Hours Numb er of Shutd Shutd ow ns

(hours)

Description: MTB MTBS is a me a sure ure tha t c om b ine s the effe c ts of inherent m a c hine reliability and the effectiveness of the equipment management orga niza niza tion in in its its a b ility to influe influe nc e results results throug h p rob lem a vo id a nc e. Da ta So urce (s): (s): Op erating erating hours hours ob ta ined from from ma c hine hine servi servic c e mete r rea d ing . Note, hours obtained from dispatch systems frequently do not agree with ma c hine hine SMU SMU d ue to c od ing of p rod uc tion tion d elays elays,, etc . Note that hours hours ta ken from from ma c hine hine SMU will will be highe highe r tha n tho se ta ken from from d isp a tc h, oftentimes b y a s muc h a s 10 p erce nt. Note: Production delay hours may not be tracked and accounted for sep a ra tely a nd a re the refore inc inc lud ed in the to ta l op erating hours. hours. Sites that use dispatch systems may track and code production delay hours separate from operating hours hence they must be acquired from dispatch. Shutd ow n co unt ob ta ined from from ma c hine hine wo rkord kord er his history a nd d isp a tc h system ystem . Dis Disp a tc h infor informa ma tion tion must must be used used to a c c ount for shutd ow n events that a re not ac c om pa nied nied by a w ork orkord ord er. er. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. Caterpillar. All Rights Reserved. Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Ben c hm a rks & Ta rg et s: MTBS b en c hm a rks va ry sig nific a ntly b y ma c hine m o d el, the ir rela tive size , a g e a nd d esig n “ ma turity” a nd c om p lexity. MTBS for la rge Off Hig hwa y Truc ks in the 785 – 793 size c la ss is ve ry w e ll d o c um en te d . The b e nc hm a rk for a flee t of ne w truc ks is 80 hours; that o f a “ ma ture” flee t (one tha t ha s und ergo ne its first round of ma jor co mp one nt reb uild s) is 60 hours. Sinc e b y d efinition these b enc hma rks rep resent d oc ume nted , be st-inclass performance sustainable over time, we are frequently asked to a ssess p erforma nc e throug h a rang e o f results. The fo llow ing ta b le rep resent s ou r be st jud g me nt in this a rea .

MTBS 50 to 60 hours 40 to 50 hours 30 to 40 hours 20 to 30 hours < 20 hours

Assessm ent / Cha rac teristic s Exce llent; high % of sche duled dow ntime; Equipm ent Mg mt. orga nization is highly proactive. Acc epta ble; ma jority of do wntime is sche duled ; substantial em phasis on Equipm ent Mgm t. Marginal; ap prox. half of all d owntime is sche duled; Equipm ent Mg mt. disciplines not fully functional. Fair; < 40% downtime is scheduled; minimal effort on Equipment Mgmt. Poor; only PM’s are sch ed uled ; Equip me nt Mg mt. orga nization is purely reac tive.

Ta b le 1: Site p erforma nc e t hrough ra ng e o f MTBS

Benchmarks for trucks smaller than the 785 and the 797 are less well kno w n a ltho ug h it is b e lieve d tha t M TBS fo r truc ks in the 769 – 777 size c lass will be significantly higher (as much 30 to 40%) while that of the 797 will be p e rha p s 10% lo w er. Simila rly, b enc hm a rks for othe r la rg e m ining eq uipm en t a re no t w ell d oc ume nted . How eve r, ind ic a tions a re tha t onc e MTBS d a ta is c ollec ted , a na lyzed a nd va lida te d , the results will fa ll into the follow ing ra ng es: Mac hine / Mod el

MTBS

D10 / D11 TTT’s

55 to 75 ho urs

992 / 994 WL’s

55 to 75 ho urs

16 MG

95 to 105 hours

24 MG

55 to 75 hours

5000 HEX

55 to 75 ho urs

Ta b le 2: MTBS g uid elines for mining m a c hines Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

1.1.2 - Mea n Tim e To Rep a ir (MTTR) Definition: The a verag e d ow ntime for ma c hine stop pa ge s … the a verag e d ura tion of e q uipm ent d ow ntime e ven ts, exp ressed in hou rs. Ca lc ulation Method ology:

MTTR =

To ta l Dow ntime Hours

(hours)

Numb er of Shutd ow ns

Description: Me a n Time To Rep a ir (MTTR) is a p e rfo rm a nc e m ea sure tha t q ua ntifies repair turnaround time, i.e. how quickly (or slowly) a machine is returned to servic e o nc e a d ow ntime inc id ent o c c urs. MTTR c om b ines the e ffec ts of inherent machine maintainability / serviceability and the efficiency of the equipment management organization in delivering rapid remedial a c tion in the exec ution o f need ed rep a irs.

Da ta So urce (s): Downtime hours obtained from machine workorder history and dispatch system . Disp a tc h informa tion must be used to a c c ount for d ow ntime tha t is not a c c om p a nied b y a w orkorde r. It is essential to no te tha t rep a ir d elay time should b e inc lud ed in the do wntime history c a lc ula tion. If d elay times a re know n, MTTR should b e c a lc ula ted b oth with a nd without delays. Shutd ow n co unt ob ta ined from m ac hine wo rkord er history a nd d isp at c h system . Onc e again, dispatch information must be used to a c c ount for shutdo wn e vents that

are not ac c omp anied by a workorde r. Ben c hm a rks & Ta rg et s: MTTR b enc hm a rks va ry som ew ha t b y m a c hine m od el, their rela tive size a nd d esig n c om p lexity b ut to a mu c h lesser extent tha n MTBS; ma c hine a g e is the p rima ry d river o f MTTR. MTTR fo r la rg e Off Hig hw a y Truc ks in the 785 – 793 size c lass is ve ry we ll d o c um e nte d . The b e nc hm a rk fo r a flee t of truc ks in th e 785 – 793 size c lass is 3 to 6 ho urs. MTTR fo r new truc ks sho uld Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

be close to the low end of the range while that of a “mature” fleet (one that has undergone its first round of major component rebuilds) should be c loser to t he hig h end o f the ra ng e . This is a result o f the re la tive c om p lexity of the rep a irs see n o n ne w versus “ ma ture” ma c hines. Benchmarks for trucks smaller than the 785 and the 797 are less well kno w n a ltho ug h it is b elieve d tha t M TTR fo r truc ks in the 769 – 777 size c lass will be slightly lower (10 to 20%) while that of the 797 will be perhaps 10% higher. Simila rly, b enc hm a rks for othe r la rg e m ining eq uipm en t a re no t w ell d oc ume nted . How eve r, ind ic a tions a re tha t onc e MTTR d a ta is c ollec ted , analyzed and validated, the results will fall into much the same range as larg e O HT flee ts w ith large r m a c hine s, e.g . 24H MG a nd 5000 se ries HEX, b eing a s mu c h a s 30 to 40% highe r.

1.1.3 - Perc enta g e Sc hed uled Dow ntime Definition: The p erce nta g e of tota l d ow ntime ho urs p erformed in a g iven p eriod tha t have be en planned and sc heduled . Ca lc ulation Method ology:

% Sc hed uled Dow ntim e =

Sc hed uled Dow ntime Hours To ta l Dow ntime Hours

X 100

(%)

Description: A high percentage of unscheduled downtime incidents results in very inefficient use of resources and excessive costs since personnel are frequently shuffled from job to job and facilities and manpower req uireme nts need to b e suffic iently large to a c c om mo da te hug e swings in the numb er of ma c hines d ow n for rep a irs. Data c ollec ted from m ine studies has shown that the average downtime for unplanned / unscheduled work is up to eight times greater than the downtime for planned / sc hed uled a c tivity.

Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Da ta So urce (s): Downtime hours obtained from machine workorder history and dispatch system . Disp a tc h informa tion must be used to a c c ount for d ow ntime tha t is not a c c om p a nied b y a w orkorde r. It is essential to no te tha t rep a ir d elay time should b e include d in the d ow ntime history c a lc ula tion. Ind ivid ua l wo rkord ers should b e c od ed a s “ sc hed uled ” or “unsc hed uled in order to tra c k the num b er of d ow ntime hours tha t are sc hed uled . Ben c hm a rks & Ta rg et s: % Sc he d uled Dow ntime Ho urs fo r la rg e Off Hig hw a y Truc ks in the 785 – 793 size class is very well documented. Mines with highly effective equipment management processes in place are able to execute 80% of its ma intena nc e and rep a ir d ow ntime a c tivity on a sc hed uled b a sis. We believe that this criterion holds true for other mining equipment as well however requirements for less utilized, non-production equipment may be so m e w ha t less.

1.1.4 - Co ntra c tual Ava ilab ility Definition: The ra tio o f time tha t a m a c hine is c a p a b le o f func tioning in the intend ed operation (available hours) to total hours in the period under consideration (typically either based on calendar or scheduled hours), exp ressed a s a p erc en ta g e. The c a lc ula tion of a va ila b le hours is no t a p ure c a lc ula tion sinc e the result is a me nd ed b y d ow ntime hours tha t a re sp ec ific a lly exclude d or limited b y the te rms of the c ontrac t. Ca lc ulation Method ology:

Tota l Ho urs * - MARC Dow ntime Ho urs Contractual Availability = To ta l Hours *

X 100

*NOTE: “ Tot a l Hours” a re typ ic a lly eithe r b a sed on c a lend a r or sc hed uled hours depending upon the calculation methodology defined in the a g ree me nt or c ustom er expe c ta tions. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

(%)

Description: Contracts are written largely to ensure that production equipment is a va ila b le for op eration a suffic ient num b er of hours to e na b le the mine to me et its p rod uc tion g oa ls a t a rea sona b le, p red ete rmined o p erating c ost. The sp ec ific p rov isions of a c on tra c tua l a va ila b ility g ua ra nte e va ry significantly from site to site, i.e. time that the contractor will be given credit for (available hours), time that the contractor will be held accountable for (contractual downtime), as well as specific exclusions, e.g. tires, dump b od ies, we ld ing , etc . a re d efined in d eta il in the c ontrac t. Furthermore, contracts frequently specify caps or limits on downtime that apply to things such as delays waiting on facilities, repair equipment and or other support infrastructure that the contractor is not expected to p rov id e an d ha s little c on trol ov er. Bec a use these exc lusions a nd limitations vary so widely from one site to the next, it is not possible to link p erforma nc e in this a rea to a ny kind of Benc hma rk nor d oe s it ma ke a ny sense to a ttem p t to ma ke c om p a risons from one site to the next. Ta rg et p erforma nc e should b e c om p lia nc e w ith the p rovisions d efined within the c ontra c t or, in the a bsenc e of a c ontra c t, with c ustome r expec tations. Da ta So urce (s): Tota l c a lend a r hours is eq ua l to the tota l time in the p eriod to b e analyzed, e.g. 8760 hours / year, 720 hours / 30 day month, 168 hours / wee k, etc. If the available hours calculation involves the combination of operating hours, stand-by hours, production delay hours and operational delay hours (as it does in many instances), that information can be obtained from the machine service meter reading and information coded within the d isp a tc h syste m. MARC d ow ntime hours a re o b ta ined from the m a c hine w orkorder history a s we ll a s the d isp a tc h system . Disp a tc h informa tion must be used to ac c ount for d ow ntime tha t is not ac c om pa nied b y a w orkord er. It is essential that the machine repair history contain detail sufficient to determine if individual downtime events are excluded from the MARC downtime calculation. Ben c hm a rks & Ta rg et s: There is no Ben c hm a rk tha t is a p p lica b le to the C on tra c tua l Ava ila b ility pe rforma nc e me tric . Ta rge t pe rforma nc e should b e c om plia nc e with th e Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

provisions defined within the contract or, in the absence of a contract, with c ustom er exp ec ta tions.

1.1.5 - Ava ilab ility Inde x Definition: The ra tio o f MTBS (a ve ra g e shutd o w n freq ue nc y) to the sum o f MTBS a nd MTTR (a ve ra g e shutd ow n d ura tion), exp ressed a s a p erc ent a g e.

Ca lc ulation Method ology:

Ava ila b ility Ind ex =

MTBS X 100

(%)

MTBS + MTTR Description: Ava ila bility is the result of the freq uenc y a nd d ura tion of d ow ntime eve nts (shutd ow ns). Bec a use o f the m a the ma tica l relat ionship b et we en M TBS, MTTR a nd Ava ila b ility Ind ex, the result show s wh ich of the ot he r two fa c to rs ha d the g rea te st influe nc e up on th a t result. This a llow s ma na g em en t to react appropriately to changes in the Availability Index and by focusing its effo rt a nd re so urc e s o n t he fre q ue nc y (M TBS) o r du ra tion (MTTR) o f do wntime events. Da ta So urce (s): Sinc e Ava ila b ility Ind e x is d erive d fro m MTBS a nd MTTR, the d a ta so urc es for those two me tric s a re a p p lic a b le here a s we ll. Ben c hm a rks & Ta rg et s: Availability Index benchmarks vary significantly by machine model, their relative size, ag e a nd d esign “ ma turity” a nd c om p lexity. Ava ila b ility Ind ex fo r la rg e Off Hig hw a y Truc ks in th e 785 – 793 size c lass is ve ry we ll d oc ume nted . The b enc hma rk for a flee t of new truc ks 92%; tha t of a “ ma ture” flee t (one tha t has und ergone its first round of ma jor c om p one nt reb uild s) is 88%. Benchmarks for truck smaller than the 785 and the 797 are less well known a ltho ug h it is b elieve d tha t the Ava ila b ility Ind ex for truc ks in the 769 – 777 Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

size class will be somewhat higher (possibly 2 to 3%) while that of the 797 will be p erha p s 1 to 2% low er. Simila rly, b enc hm a rks for othe r la rg e m ining eq uipm en t a re no t w ell do c umented . How ever, indica tions are that onc e the da ta is c ollec ted , analyzed and validated, the results will fall into much the same range as larg e O HT flee ts w ith large r m a c hine s, e.g . 24H MG a nd 5000 se ries HEX, being as much as 3 to 4% lower and smaller machines, e.g. 16H, being 1 o r 2% hig he r.

1.1.6 - Ma intenanc e Ra tio Definition: Maintenance Ratio is a dimensionless ratio of maintenance and repair ma n-hours to ma c hine o p erating hours. Ca lc ulation Method ology:

Ma intena nc e & Rep a ir Ma n-Hours

Ma intena nc e Ra tio

c ha r e d

=

Op erating Hours

Description: Maintenance Ratio is an indication of the amount of effort required to keep equipment in service as well as the efficiency with which labor is d ep loye d a nd the effec tivene ss of the wo rkforce in c a rrying out its d uties. Maintenanc e Ratio c an b e c alculated as either “ c harge d” or “ direc t”. “Charged” Maintenance Ratio considers only workorder man-hours (direc t la b or). Rep a ir sho p , e.g . Co mp on en t Reb uild C en ter, la b or is no t inc lud ed in the ca lc ula tion. “ Ove ra ll” Ma intena nc e Ra tio inc lud es a ll the eleme nts of “ c ha rg ed ” Ma intena nc e Ra tio p lus sta ff, sup ervision a nd id le time. Da ta So urce (s): Maintenance and repair man-hours are obtained from the work order histo ry. The result sho uld inc lude a c tua l time sp e nt w o rking o n a ll fo rm s o f ma inte na nc e, rep a irs a nd mo d ific a tion s a s we ll a s ineffic ienc ies tha t result from inspection and diagnostic time or any delay or wait time for bay space, parts, tooling, literature, repair support equipment, decision ma king, etc. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Operating hours are obtained from machine service meter reading and onc e a g a in should inc lud e p rod uc tion d elay hours. Note, hours ob ta ined from d isp a tc h system s freq uently do not a g ree with ma c hine SMR d ue to c od ing o f prod uc tion d elays, etc . Ben c hm a rks & Ta rg et s: Maintenance Ratio benchmarks vary significantly by machine model, their relative size, age and design “maturity” and complexity. Ma inten a nc e Ra tio fo r large Off Hig hw a y Truc ks in the 785 – 793 size c lass is ve ry we ll d oc um ent ed . The b enc hm a rk for a flee t of ne w truc ks is 0.20 man-hours/ operating hour; that of a “mature” fleet (one that has und ergo ne its first round of ma jor c om p on ent reb uild s) is 0.30 ma n- ho urs/ operating hour. Sinc e b y d efinition these b enc hma rks rep resent d oc ume nted , be st-inclass performance sustainable over time, we are frequently asked to a ssess p erforma nc e throug h a ran g e o f results. The fo llow ing (ta b le 3) rep resent s ou r be st jud g me nt in this a rea . MR

Assessm ent / Cha rac teristic s

0.30 to 0.35

Exce llent; high % of sche duled downtime ; Equipm ent Mg mt. orga nization is highly proactive.

0.35 to 0.40

Ac ce ptab le; ma jority of dow ntime is sched uled; substantial em phasis on Equipm ent Mgm t.

0.40 to 0.50

Marginal; ap prox. half of all d owntime is sche duled; Equipm ent Mg mt. disciplines not fully functional.

0.50 to 0.60

Fair; < 40% do wntime is sch ed uled ; minim al effort on Equip me nt Mg mt.

> 0.60

Poor; only PM’s are sche dule d; Equip me nt Mg mt. orga nization is purely reac tive.

Tab le 3: Site performa nc e through rang e of Ma intena nc e Ratios

Benchmarks for trucks smaller than the 785 and the 797 are less well known although it is believed that Maintenance Ratio for trucks in the 769 – 777 size c la ss w ill be sligh tly low e r w hile th a t o f th e 797 w ill be som ew ha t higher. Simila rly, b enc hm a rks for othe r la rg e m ining eq uipm en t a re no t w ell do c umented . How ever, indica tions are that onc e Ma intenanc e Ratio data is collected, analyzed and validated, the results will fall into the ra nge s show n in the tab le b elow. It is imp ortant to note he re tha t ma c hine a p p lica tion w ill p la y a role in Ma inte na nc e Ra tio. This is p a rticula rly true in the c a se o f la rg e Tra c k-typ e Tra c to rs tha t c a n b e d ep loyed a s either p rod uc tion or sup p ort eq uip me nt. (Refer to ta b le 4 below). Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Mac hine / Model

MR

D10 / D11 TTT’s

0.40 to 0.50

992 / 994 WL’s

0.35 to 0.45

16 MG

0.10 to 0.15

24 MG

0.15 to 0.20

5000 HEX

0.50 to 0.60

Tab le 4: Ma intena nc e Rat io guidelines for mining m a c hines

1.1.7 - Top Prob lems Sum m a ry Definition: The d istribution o f p rob lem s a ffec ting a flee t o f eq uip me nt ra nked in te rms o f MTBS, MTTR, im p a c t o n A va ila b ility a nd Co sts. Ca lc ulation Method ology:

MTBS (b y syste m) =

MTTR (b y syste m ) =

Imp a c t on Ava ila b ility

(by system)

Op erating Hours

(hours)

Numb er of Shutd ow ns (b y syste m)

Downtime Hours (b y syste m)

(hours)

Numb er of Shutd ow ns (by syste m)

= ( 1 – Ava ila b ility

(total mac hine)

Cost

Co st p er Hour (b y syste m) =

)

Downtime Hours (by system) X

(%) To ta l Dow ntime Ho urs (machine)

(b y syste m)

(US & / ho ur)

Op erating Hours


Description: All m ining sup p o rt op era tions ha ve lim ited reso urc es. The m o st suc c e ssful operations are those that have a clear understanding of the problems and issues facing them and are thus in a position to establish priorities in order to focus their efforts and allocate the appropriate resources on rem ed ia l or c on ta inmen t stra teg ies throug h c on tinuo us imp rov em en t. The identification and quantification of top problems by component (e.g. engine, transmission, …), system (e.g. hydraulics, electrical, …) or even process (e.g. PM) facilitates the understanding of the extent that each area is having an influence on various criteria that comprise the success of a mining sup p ort op eration , i.e. shutd ow n freq ue nc y (MTBS), shutd ow n d urat ion (MTTR), imp a c t on Ava ila b ility a nd C o sts. With this kno w led g e the Projec t Ma na g er is a b le to “ d rill d ow n” to the key issues fa c ing his site and apply the necessary resources in the most efficient manner to imp ro ve his situa tion. Da ta So urce (s): Operating hours are obtained from machine service meter reading. Note, hours ob ta ined from d isp a tc h system s freq uently do not a g ree with ma c hine SMR due to c od ing of p rod uction d elays, etc. Shutd ow n c ount is ob ta ined from ma c hine w orkorder history and d isp a tc h system . Disp a tc h informa tion must be used to a c c ount for shutd ow n eve nts tha t are not a c c ounte d for b y a w orkorde r. Shutd ow n co unt must b e d ete rmined ind ivid ua lly for ea c h a rea of the m a c hine a s we ll a s for the machine as a whole in order to assess not only the contribution of each a rea b ut a lso to c a lc ula te Ava ila bility Ind ex. Downtime hours obtained from machine workorder history and dispatch system . Disp a tc h informa tion must be used to a c c ount for d ow ntime tha t is not a c c om p a nied b y a w orkorde r. It is essential to no te tha t rep a ir d elay time s should b e include d in the d ow ntime history c a lc ula tion. If d elay times a re know n, MTTR should b e c a lc ula ted b oth with a nd without d ela ys. As is the c a se with shutd ow n c ou nt, d ow ntime must b e determined individually for each area of the machine as well as the ma c hine a s a who le in order to a ssess the c ontribution o f ea c h a rea . Tota l c ost to sup p ort and ma inta in ea c h of the system s a nd c om p one nts on the m a c hine. At a m inimum it is vita l to know the b rea kdo wn for c osts of rep a irs a nd reb uild s of ea c h major c om p one nt on the ma c hine. Mo st recordskeeping systems we have studied do a fairly poor job of Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

documenting costs but if Project Management is to have any opportunity to m a na g e c ontrac t p rofitab ility, c osts must be known. Ben c hm a rks & Ta rg et s: There is no set of Benc hma rks tha t is a p p lica b le t o t his me tric . How ev er, ove r the c ourse of investiga tion d uring EMR’ s we d eve lop ed a c ollec tion of g ene ric referenc e g uid eline s for la rg e Off Hig hw a y Truc ks in the 785 – 793 size c la ss tha t c a n b e use d a s a g a ug e to ev a lua te MTBS, MTTR a nd imp a c t on Ava ila bility. This referenc e d efines wha t we b elieve to b e a reasonable level of acceptability for frequency of downtime events (MTBS), d urat ion o f do w ntime ev en ts (MTTR) an d im p a c t on A va ila b ility for ea c h of the major areas on the ma c hine. The d a ta is rep resent a tive o f a site o p erating a t a n Ava ila b ility Ind ex of a p p roxima te ly 90% a nd is, of c ou rse, g en eric sinc e a c tua l results a c hieved at any given mine are site-specific because results of this kind are a function of not only application severity but also of the operating environment, the maintenance the equipment receives and product design shortcomings that are particular to machines either by model or within a g ive n ra ng e o f seria l num b ers. The “ Ge neric Pa ret o Refe ren c e – La rg e Off Hig hw a y Truc ks” inc lud ed in the Appendix can be used as a baseline until Project Management has sufficient individual site-specific experience and history to determine how this referenc e c a n be mo d ified to fit the a p p lic a tion in question. Sinc e there are ma ny fac tors other tha n eq uipm ent ma nag eme nt that influence costs (labor rates, transportation costs, import duties, taxes, etc .), it is imp ossib le to d efine Ben c hm a rks tha t a re universa lly a p p lic a b le to a ny given ma c hine mo d el. This be ing the c ase, we rec om mend that bud ge tary c ost and c om p onent life projec tions be used to d efine ta rge t c ost pe r hour figures and that a c tual cost pe rforma nc e b e c om pa red to those targets in order to determine if any particular area is out of line with expectations.

1.1.8 - Asset Utilization Definition: The p rop ortion o f time tha t a ma c hine is op erating (op erating hours) divided by the total calendar time in the period, expressed as a percentage. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Ca lc ulation Method ology:

Op erating Hours

Asset Utilization =

X 100

(%)

Tot a l Ca len d a r Ho urs

Description: How effectively the Operations Department schedules equipment and efficiently it utilizes that equipment has significant implications for Ma inte na nc e. If ma c hines a re sc he d uled for use 24 ho urs a d a y, 7 d a ys a week, Maintenance must respond by working with Operations to find windows of opportunity in which maintenance and repairs can be p erforme d without increa sing do wn time. These op p ortunities typ ic a lly occur during scheduled shutdowns but they may also come at shift c ha ng e, lunc h b rea ks or during o p erationa l d ela ys suc h a s d uring b la sting or fueling of equipment. In all circumstances, Operations and Maintenance need to recognize that they are working together toward common goals … high availability, good machine reliability and the low est p ossib le c ost p er unit of p rod uc tion. Da ta So urce (s): Operating hours are obtained from machine service meter reading and should inc lud e prod uc tion de la y hours. Note, hours ob ta ined from d isp a tc h system s freq uently do not a gree with ma c hine SMU d ue to c od ing of prod uc tion de la ys, etc . Note that hours ta ken from m a c hine SMU will b e hig her tha n tho se ta ken from d isp a tc h, often time s b y a s mu c h a s 10 p erce nt. Tota l c a lend a r hours is eq ua l to the tota l time in the p eriod to b e analyzed, e.g. 8760 hours / year, 720 hours / 30 day month, 168 hours / we ek, etc . Ben c hm a rks & Ta rg et s: Asse t Utiliza tio n fo r large Off Hig hw a y Truc ks in the 785 – 793 size c lass is very well documented. Mines with highly effective equipment management processes in place are able to achieve Asset Utilization of 90%, ove r 7800 op erating ho urs p er yea r. We b eliev e th a t this Ben c hm a rk is valid for other production mining equipment however the Benchmark for less utilized, non-production equipment, although unknown, may be signific a nt ly less. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

1.1.9 - PIP / PSP Co mp letio n Ra te Definition: A tracking tool used to monitor the status of implementation of factory programs. Ca lc ulation Method ology: Fa c tory p rog ra m c om p letion sta tus is c a lc ula ted a s the ratio o f prog ra ms completed on a machine-by-machine basis relative to the number of program s that a re a c tive a nd a p plic a ble at the time unde r c onside ra tion. This ra tio sho uld b e exp ressed a s a p erc en ta g e. Prog ra ms tha t are d efined a s "a fter fa ilure" should no t b e inc lud ed in the c a lc ula tion. Da ta So urce (s): Fa c tory prog ra ms a re rec eived on site via the d ea ler Tec hnic a l Communications staff and include all of the information necessary to d ete rmine a p p lic a b ility a nd mo nitor their co mp letion sta tus, i.e. p rog ra m identification number, dates of issue and termination, and program type. Machine serial number and hourmeter information obtained from the ma c hine history at the site . Ben c hm a rks & Ta rg et s: Sinc e fa c tors suc h a s p a rts a va ila b ility c a n imp a c t on m a na g em ent's ab ility to c om p lete a program and in some c a ses p rog ra m e xec ution c a n be delayed to coincide with other related work (which may be a valid decision on the part of management), there is no Benchmark that is a p p lic a b le to this me tric. How ev er, c om p lia nc e with this d isc ip line is viewed as critical to the success of a project and common sense would dicta te that a highe r pe rc entag e of co mp letion of outstand ing p rog ra ms is d esira b le. Clea rly, no p rog ra m sho uld b e p ermitted to run b eyo nd its termination date without being addressed unless it is an after failure only program.

2.1 Preventive Ma intena nc e The fo llow ing me tric s a re indica tors of the p erforma nc e a nd / or contributions of the Preventive Maintenance process to the end results of the p rojec t, i.e. eq uip me nt relia bility a nd a va ila b ility. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

2.1.1 - MTBS a fter PM

MTBS after PM

Tota l Op era ting Hours to First Sto p

=

(hours)

Numb er of PM Servic es

The a ve ra g e op erat ing ho urs to the first sto p a fter ea c h PM servic e is a va lid ind ic a tion o f PM q ua lity an d e ffec tivene ss. The Ben c hm a rk (b est in c lass) fo r large Off Hig hw a y Truc ks is 105 ho urs; a re a listic ta rge t is 2 to 3 time s the o ve ra ll MTBS. Tra c king a nd trend ing th is m e tric m o nth ly offe rs a rea son a b le rep resent a tion o f PM q ua lity an d e ffec tive ness.

2.1.2 - Una va ilab ility PM

Unavailability

PM

=

(1 - Ava ila b ility) X

PM Dow ntime Ho urs Tota l Dow ntime Ho urs

X 100

Una va ila bility d ue to Preve ntive M a intena nc e q ua ntifies the imp a c t of PM on a va ila b ility. There is no Ben c hm a rk a ssoc ia ted with this me tric b ut a reasonably valid target is in the range of 2.75 to 3.25% for large Off Hig hw a y Truc ks in the 785 – 793 size c lass. The re sult of t his m ea sure sho uld b e t a ken in c o nte xt with M TBS after PM . If una va ila b ility d ue t o PM is b elow the ran g e a nd MTBS after PM is low, it is a safe assumption that insufficient time and effort is being placed on PM. Co nve rsely, if una va ila b ility d ue t o PM is a b ov e t he ran g e a nd MTBS aft er PM is high, one can assume that the site is placing substantial emphasis on the va lue o f PM.

2.1.3 - MTTR PM

MTTR PM =

Tota l PM Dow ntime Hours

(hours)

Numb er of PM Servic es

The a verag e d ow ntime hours d ed ic a ted to PM is a n ind ic a tion o f PM efficienc y. There is no Ben c hm a rk a ssoc ia ted with this me a sure b ut a reasonable target is in the 7.75 to 8.5 hour range for large Off Highway Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

(%)

Truc ks tha t are on a 250 ho ur PM se rvic e interva l. The ta rg et fo r equipment that is on a 500 hour PM service interval is higher, i.e. double tha t of the 250 hour interva l. Onc e a ga in, if ave ra ge PM d ow ntime is the ra ng e, it ma y ind ic a te a nd MTBS after PM is low, it is a safe assumption that insufficient time and effort is being placed on PM. Conversely, if average PM d ow ntime is a b ove the rang e a nd MTBSaft er PM is high, one c a n a ssum e that the site is placing substantial emphasis on the value of PM. Ad d itiona lly, one should c onsid er the imp a c t of e ffic ienc y fac tors suc h a s facilities, tooling, training, planning & scheduling, etc. when assessing MTTR PM.

2.1.4 - Service Accuracy A measurement of Preventive Maintenance execution timeliness based on a statistical calculation that predicts the probability that the next PM service will occur within the recommended range (+/- 25 hours of target interval). The c a lc ula tion is b a sed up on p a st p erforma nc e a nd a ssume s tha t PM interva ls a re no rma lly d istribute d a b ou t the me a n. The Benc hm a rk fo r S.A. is 95% b ut a n a g g ressive ta rg et th a t w ill yield exc ellent results is 90%.

2.1.5 - Ba c klo g s exec uted d uring PM Backlogs executed during Preventive Maintenance is a good indication of how well the organization is using the “window of opportunity” presented by PM to maintain the equipment at a standard that will enh a nc e p rod uc t relia b ly. There is no Ben c hm a rk or ta rg et fo r this measure.

2.1.6 - Bac klog s ge nerated during PM The num b er of de fec ts id entified a nd ente red into the Ba c klog Management system during the execution of Preventive Maintenance. Sinc e this me a sure is a d irec t func tion of the num b er of ma c hines b eing monitored as well as their condition, no Benchmarks or targets are applicable. Backlogs generated during PM quantifies the use of the “window of opportunity” presented during the PM shutdown for defect de tec tion (an e leme nt of Cond ition Monitoring).


2.2 Co nd ition Monitoring The fo llow ing me tric s a re indica tors of the p erforma nc e a nd / or contributions of the Condition Monitoring process to the end results of the project.

2.2.1 - Me a n Tim e Between Fa ilure s (MTBF)

MTBF (Me a n Time Be tw e en Fa ilures)

=

Operating Hours

(hours)

Numb er of Fa ilures

The a verag e o p erating time b etw ee n eq uip me nt failures; the inverse o f failure frequency, expressed in hours. Failures may be the result of technical product issues, i.e. equipment unreliability, or due to maintenance / repair neglect, i.e. equipment management ineffec tivene ss in the a rea of p rob lem av oida nc e. We ha ve not esta b lished a Benc hm a rk for MTBF a nd d o no t ha ve suffic ient c on fid en c e at this time to provide a reasonable target. (Please see the Glossary for our d efinition of e q uip me nt Fa ilure.)

2.2.2 - Una va ilab ility

Unavailability

U/S

=

(1

Unscheduled

Availability) X

Unsc hed uled Dow ntime Tot a l Dow ntime

Unavailability due to unscheduled downtime quantifies the impact of unsc hed uled ev en ts on a va ila b ility. There is no Ben c hm a rk assoc ia ted with this me tric b ut a rea son a b ly va lid ta rg et is < 2% for la rg e O ff Hig hw a y Truc ks in the 785 – 793 size c lass. If una va ila b ility d ue to unsc he d uled downtime is significantly higher than 2%, it is reasonable to assume that gaps exist in the detect-plan-execute cycle therefore improvements to the Co nd ition Mo nitoring , Pla nning & Sc hed uling a nd / or rep a ir exec ution areas will be necessary. Increasing unavailability due to unscheduled downtime is a valid predictor of pending problems and may very well p red ic t future shortag es of m a np ow er and fac ilities.


X 100 (%)

2.2.3 - Failure Reduction

Fa ilure Red uc tion (FR)

=

Unsc hed uled hours (6 m o nth s RA)

_ Unsc hed uled hou rs (last month)

Unsc hed uled ho urs (6 month rolling a vera ge )

X 100 (%)

Failure Reduction is a means of quantifying the impact of Condition Mo nitoring in its effo rts to wa rd fa ilure / p rob lem a vo id a nc e. Since unscheduled events are inherently more difficult and inefficient to deal with in terms of the time required to make unscheduled (unplanned) repairs, Failure Reduction should be the primary focus of Condition Monitoring activities. Because the opportunity to improve in this area is highly dependent upon the amount of unscheduled downtime taking p la c e a t the site, the re is no Benc hm a rk or targe t for Fa ilure Red uc tion. In any event, the result should be positive indicating a decline in the percentag e of unsc heduled do wntime.

2.2.4 - Co nd ition Mo nitoring Tota l Sa ving s

CM Total Sa vin s =

CM Cost Sa vin

C M Pro ra m C o st

Co nd ition Mo nitoring Tot a l Sa vings d efines the “ Va lue Prop osition ” for Condition Monitoring. In other words, the total savings generated by Condition Monitoring (cost of after-failure repairs – cost of preventive, before-failure repairs) must be greater than the cost of implementation a nd exec ution of the C ond ition M onito ring p rog ra m. There is no Benchmark for this metric but the target should be a positive value (net savings as a result of Condition Monitoring).

2.2.5 - Tota l Ba c klo g s Ge nerated

Ba c klog s Ge nerate d = To ta l Ba c klo g s Ge ne ra te d in th e Perio d (To ta l) The num b er of de fec ts id entified a nd ente red into the Ba c klog Ma na ge me nt system d uring a sp ec ified p eriod (typica lly one mo nth). This metric assesses the Condition Monitoring effort in and ability to suc c essfully d ete c t p ot en tial prob lem s b efo re fa ilure. Sinc e th is me a sure is Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

US$

a d irec t function o f the numb er of ma c hines b eing m onitored , there is no Benc hma rk or ta rg et.

2.2.6 - Working on Ta rge t

Working o n Ta rg et = % Ba c klo g s o n Prob lem List The p erce nta ge of Ba c klog s g ene ra ted whic h a d d ress issues tha t a p p ea r o n th e “ To p 10” histo ric a l prob lem list. The result yie ld s the % o f Co nd itio n Mo nitoring a c tions tha t a re “ On Ta rg et” relat ive t o the key issues a ffec ting site p erforma nc e. There is no Ben c hm a rk or targe t fo r this me tric ho w ev er, if all issues on the problem list are not producing Backlogs, the Condition Mo nitoring effort ma y be misd irec ted .

2.2.7 - Ba c klog s Gene ra ted b y Origin

Ba c klog s Ge nerate d

=

Ba c klog s Ge nerate d in the Period b y a rea of o rig in

(b y O rig in) Backlogs Generated by Origin identifies which areas that are or are not c on trib uting to effo rts b y Co nd ition M onito ring in fa ilure d et ec tion. There is no Benc hma rk for Ba c klog s Ge nerated b y Origin, how eve r, if the q ua ntity of Backlogs generated by operators, inspectors, the PM crew, the shop crew, etc is low, additional emphasis should be placed on the offending party(s) to encourage their participation in the defect detection process. Conversely, if the percentage of “shop-found” defects is d isp rop ortiona tely high, the o ther area s must b e e nc ourag ed to inc rea se their involvement since “shop found” defects are typically far less effic iently exec uted d ue to the inab ility to p la n the w orkloa d . There are no Benc hm a rks or ta rg et s relat ed to this me tric.

2.2.8 - Dete c tion Level

Potential Failure Detec tion (PFD)

Rec orde d Ba c klog s =

X 100 (%)

Tot a l d efe c ts p end ing


This me tric is b a sed on a c om p a rison b etw ee n the numb er of Ba c klog s recorded in the system and the defects that can be detected in an insp ec tion of a ra nd om ly selec ted sa mp le o f ma c hines (10% of to ta l flee t a t minimu m). Althoug h insp ec tions a re limited to visua l insp ec tion, the Potential Failure Detection level can be used to assess the level of detection of Condition Monitoring.

2.3 Ba c klog Ma nag em ent The fo llow ing me tric s a re used to e va lua te the a b ility of the Ba c klog Management process to prioritize, control and manage problems identified through Condition Monitoring such that they do not result in unnec essa ry d ow ntime .

2.3.1 - Tota l Ba c klog s Pend ing The to ta l numb er of de fec ts id entified b y Cond ition M onitoring a nd pending in the Backlog Management process. An indication of the p end ing wo rkloa d a nd risk for fa ilure. Sinc e t his num b er is d ep end ent upon the size of the fleet being managed, there is no Benchmark or ta rg et for this me tric.

2.3.2 - Ba c klo g s Pend ing b y Ma c hine The to ta l numb er of d efec ts p er ma c hine identified b y Cond ition Mo nitoring a nd p end ing in the Ba c klog Ma na ge me nt p roc ess. There is no Benchmark for this metric, however a reasonable target is that there should b e no mo re tha n five p end ing Ba c klog rep a irs p er ma c hine.

2.3.3 - Tota l Ba c klo g s Ge nerated

Ba c klog s Ge nerate d = To ta l Ba c klo g s Ge ne ra te d in th e Perio d (To ta l) The num b er of de fec ts id entified a nd ente red into the Ba c klog Ma na ge me nt system d uring a sp ec ified p eriod (typica lly one mo nth). This metric assesses the Condition Monitoring effort in and ability to suc c essfully d ete c t p ot en tial prob lem s b efo re fa ilure. Sinc e th is me a sure is a d irec t function o f the numb er of ma c hines b eing m onitored , there is no Benc hma rk or ta rge t. Ba c klog ge neration should b e viewe d in the c onte xt of % sc hed uled d ow ntime a nd, if the pe rc enta ge of sc hed uled d ow ntime Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

is low , the tota l numbe r of Ba c klog s g ene ra ted should b e c orresp ond ing ly high.

2.3.4 - Tota l Ba c klo g s Exe c uted

Ba c klo g s Exec ute d

= To ta l Ba c klo g s Exec ute d in the Perio d

(To ta l) The num b er of Ba c klog rep a irs p erforme d d uring a sp ec ified p eriod (typica lly one mo nth). This me tric eva lua tes the a b ility of the ma intena nc e organization to react appropriately to correct defects identified through the C ond ition Mo nitoring p roc ess. Sinc e this me a sure is rela te d to th e number of Backlogs in the system, there is no Benchmark or target, however, there should be a balance between the number of Backlogs generated and executed.

2.3.5 - Estim a ted La b or To Rep a ir The to ta l estima ted rep a ir la b or ma n-hours req uired to e xec ute a ll of the p end ing Ba c klog s tha t have be en g ene ra ted . This me tric is a n ind ic a tion of severity of the Backlog workload and the potential availability lost if manpower resources are insufficient to accomplish the task at hand. There is no Ben c hm a rk for this me tric b ut a rea sona b le t a rg et is tha t the total estimated repair labor man-hours required to clean up the Backlog list sho uld b e < 5% of a va ila b le ma n-hours la b or for the m on th.

2.3.6 - Backlog Status Summary The Ba c klog Sta tus Summ a ry d efines the num b er of p end ing Ba c klog s that are waiting for planning (“Red phase”), waiting for parts / resources (“ Blue p ha se” ), a nd wa iting to b e e xec uted (“ Green p ha se” ). There is no Benc hm a rk or ta rg et for the Ba c klog Sta tus Summ a ry ho w ev er this me tric analyzed to identify any weak area(s) in the detect-plan-execute cycle tha t ma y be d elaying the Ba c klog rep a ir exec ution p roc ess.

2.3.7 - Ba c klo g s > 30 Da ys Old Measured from the date the Backlog was generated, this metric assesses the quality and timeliness of the response of the Backlog Management syste m in its a b ility to resp on d p roa c tively to elimina te p ot en tia l p rob lem s. It is important to note that Backlogs are potential failures, thus Backlog Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

a g e is a n ind ica tion o f the risk of fa ilure tha t a site is und e r. The re is no Benchmark for this measure however an aggressive target is that no Ba c klog s a re g rea ter than 30 d a ys old.

2.4 Planning a nd Sc hed uling The follow ing me tric s a re used to eva lua te how we ll the Pla nning a nd Sc he d uling p roc ess is orga nized a nd func tioning to e nsure tha t p la nne d activities can be accomplished both efficiently and effectively and that the y do not result in unne c essa ry do wn time.

2.4.1 - Perc enta g e Sc hed uled Dow ntime

% Sc hed uled Dow ntim e =

Sc hed uled Dow ntime Hours Tota l Dow ntime Ho urs

(%)

X 100

A high percentage of unscheduled downtime incidents results in very inefficient use of resources and excessive costs since personnel are frequently shuffled from job to job and facilities and manpower req uireme nts need to b e suffic iently large to a c c om mo da te hug e swings in the numb er of ma c hines d ow n for rep a irs. Data c ollec ted from m ine studies has shown that the average downtime for unplanned / unscheduled work is up to eight times greater than the downtime for planned / sc hed uled a c tivity. The Benc hma rk for p erce nta ge of sc hed uled d ow ntime is 80% of maintenance and repair downtime activity is executed on a scheduled b a sis. A rea so na b ly ag g re ssive t a rg et fo r m o st sites is 60%.

2.4.2 - Sc hed ule Com plianc e by Hours

Sc hed ule Co mp lia nc e = (b y ho urs)

Sc he d uled PM & Rep a ir Hours Exec ute d PM & Rep a ir Hours Sc he d uled

X 100 (%)

Sc hed ule C om p lia nc e (b y hours) is the rat io o f sc he d uled Prev ent ive Maintenance and repair downtime hours actually executed to the Preve ntive Ma intena nc e a nd rep a ir dow ntime hours sc hed uled . There is no Benc hma rk for this me tric b ut the ta rg et should b e in the ra ng e o f 90 to Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

100%. If the result is c o nsiste ntly 100%, it ma y b e a n indic a tion t ha t the schedule is too conservative (does not provide sufficient “stretch”. Co nve rsely, if the result is c on sisten tly low , it c ould me a n tha t the sc he d ule is too ambitious, the workforce is inefficient, or that the amount of unscheduled downtime during the period was such that it interfered with wo rk tha t had b ee n previously sc hed uled .

2.4.3 - Sc hed ule Com p lianc e b y Eve nts

Sc hed ule Co mp lia nc e = (by events)

Sc he d uled PM & Rep a ir Eve nts Exec ute d PM & Rep a ir Eve nts Sc he d uled

X 100

(%)

Sc hed ule C om p lia nc e (b y eve nts) is the ratio o f sc hed uled Prev ent ive Maintenance and repair events actually performed to the Preventive Maintenance and repair events scheduled. Once again, there is no Benchmark for this metric but the target should be in the range of 90 to 100%. Just a s w a s the c a se fo r Sc he d ule Co m p lianc e b y ho urs, if the result is consistently 100%, it may be an indication that the schedule is too c on serva tive (d oe s no t p rov id e suffic ient “ stretc h” . Co nve rsely, if the result is consistently low, it could mean that the schedule is too ambitious, the workforce is inefficient, or that the amount of unscheduled downtime during the period was such that it interfered with work that had been p rev iou sly sc hed uled .

2.4.4 - Com po nents Exc hang ed (sc hed uled)

Components Exchanged =

PCR’ s Sc he d uled & Exec ute d PCR’ s Exec uted

X 100

Com po nents Exc hang ed is the ratio of c om po nent rep la c eme nts scheduled and actually replaced to components replacements sc he d uled . The re is no Ben c hm a rk for this me tric b ut the ta rg et sho uld b e 100%.

2.4.5 - Estim a ted Tim e To Rep a ir Pend ing Ba c klog s The tota l estima ted rep a ir d ow ntime hours req uired to exec ute a ll of the pe nding Ba c klog s that ha ve b een ge nera ted . At the time a rep air req uest is e nte re d (Ba c klo g g e ne ra te d ) the Estima te d Time to Re p a ir (ETTR) m ust Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

(%)

be identified to permit effective planning of required corrective actions. This me tric is a n ind ica tion of extent o f the Ba c klog w orkloa d a nd the p ote ntial a va ila b ility lost if resou rc es a re insuffic ient to a c c om p lish the ta sk a t ha nd . The re is no Be nc hm a rk o r ta rg et fo r this me tric ; it se rve s a s a to o l for the Planning process to enable it to “manage” availability by sc hed uling w ork in suc h a wa y that the a vailab ility go al c a n b e m et.

2.4.6 - Estima ted Time To Rep lac e Ov erd ue Com p onents The tota l estima ted rep a ir d ow ntime hours req uired to rep la c e a ll overd ue c om p one nts. Sta nd a rd job s for c om p one nt rep la c em ent w ill forec a st the estima ted time to replac e e a c h c om p one nt. This me tric is a n ind ic a tion o f extent of the component replacement workload and the potential availability lost if resources are insufficient to accomplish those tasks. There is no Ben c hm a rk for this me tric b ut a rea sona b le t a rg et is tha t the d ow ntime req uired for c om p one nt rep la c em ent should no t result in mo re tha n 2% una va ila b ility. It is imp orta nt to no te h ere tha t sinc e flee ts ten d to come due for component replacement in “batches”, thus this metric is hig hly va ria b le a nd must b e loo ked a t o ver the long-term … 12-24 mo nths. Just a s w ith t he ETTR fo r Ba c klo g s, this m ea sure serve s a s a to o l for th e Pla nning p roc ess to e na b le it to “ ma na ge ” a va ila b ility by sc hed uling w ork in suc h a w ay tha t the ava ila bility goa l c an b e m et.

2.4.7 - Estima ted Time To Exe c ute Fa c tory Prog ram s The to ta l estima ted rep a ir d ow ntime ho urs req uired to p erform a ll ov erdue fa c to ry p ro g ra m s, i.e. PIP & PSP’ s. The p ro g ra m w ill typ ica lly de fine t he estima ted time to e xec ute e a c h p rog ra m. This me tric is a n ind ic a tion o f extent of the program execution workload and the potential availability lost if resourc e s a re insuffic ient to a c c o m p lish tho se ta sks. The re is no Benchmark for this metric but, if one assumes that program execution is relatively current, a reasonable target is that the downtime required for c om p one nt rep la c em ent should no t result in m ore tha n 1% una va ila b ility. It is important to note that since programs are generated to cover fleets of eq uipm ent, those p rog ra ms tend to c om e in “b a tc hes” , thus this me tric is highly variable and must be looked at over the long-term … 6-12 mo nths. Just a s with the ETTR for Ba c klog s a nd c om p on ent rep la c em ent , this measure serves as a tool for the Planning process to enable it to “manage” availability by scheduling work in such a way that the ava ila bility goa l c a n be m et.


2.5 Pa rts Ma na g em ent The fo llow ing me tric s a re used to d ete rmine how we ll ma intena nc e activities are supported by the parts inventory and evaluate the relationship b etw ee n the Pa rts, Pla nning & Sc hed uling a nd Ma intena nc e Dep a rtm ent s in their effo rts to a vo id unne c essa ry pa rts-relat ed d ow ntime.

2.5.1 - Warehouse Service Fill Level (instantaneous)

Servic e Fill Leve l =

Pa rts Ord ers Closed a t 1st Request Tota l Pa rts Orders

(instantaneous)

X 100

(%)

Insta nta ne ou s Servic e Fill Lev el is a p a rts ma na g em ent effic ienc y indic a tor that quantifies the percentage of individual parts requests entered against the on-site parts warehouse for repairs (including Backlog parts req uests) and filled / c losed a t the first c a ll. A reflec tion of the leve l of sa tisfa c tion of o n-site p a rts wa reh ouse p erforma nc e. The Benc hm a rk for Insta nta ne o us Servic e Fill Le ve l is 95%. An a g g re ssive ta rge t is > 90%.

2.5.2 - Servic e Fill Leve l (24 ho urs)

Servic e Fill Leve l =

Pa rts Ord ers Closed in 1st 24 Ho urs Tota l Pa rts Orders

(a fte r 24 ho urs)

X 100

(%)

Servic e Fill Lev el a fter 24 ho urs is a p a rts ma na g em ent efficienc y indic a to r that quantifies the percentage of individual parts requests entered against the on-site parts warehouse for repairs (including Backlog parts re q ue sts) an d filled / c lose d in the first 24 ho urs a fte r the first c a ll. A reflection of the level of satisfaction of on-site parts warehouse pe rforma nc e. We d o no t have suffic ient d ata to d efine a Benc hma rk but a n a g g re ssive ta rg e t is 100%.

2.5.3 - Una va ilab ility Pa rts

Unavailability

PD

=

(1 – Availability) X

Pa rts Dela y Dow ntime Tot a l Dow ntime

X 100 (%)


Unavailability due to parts delays quantifies the impact of parts delay ev e nts o n a va ila b ility. The re is no Be nc hm a rk asso c iat e d w ith this m e tric but a reasonable target is < .5%. If unavailability (downtime) due to parts d e la ys is signific a ntly hig he r tha n .5%, it m a y signify p o te ntia l p ro b lem s w ith inventory quality / quality and/or a higher than normal percentage of unp la nned d ow ntime , i.e. the ina b ility of the ma intena nc e o rga niza tion to de tec t problems in ad vanc e o f failure a nd p lan & sc hed ule the w ork a nd a ssoc iate d resou rc es a c c ording ly. If p a rts inve nto ry q ua lity / q ua lity is found to be an issue, it may be due either to the fact that the maintenance organization is not doing a good job of defining the parts inventory support requirements to the Parts Department or that the Parts Department is not delivering on its obligation to support the site with the req uired p a rts.

2.5.4 - Em erge nc y Resp onse Tim e Eme rg enc y Resp on se Time q ua ntifies the a ve ra g e resp on se time (in d a ys) to sa tisfy p a rts req ue sts tha t c a nno t b e filled insta nta ne o usly. This parameter works and should be analyzed in conjunction with Insta nta ne o us Service Fill Lev e l. The re is no Ben c hm a rk o r ta rg et fo r this metric.

2.5.5 - Pa rts Inve nto ry Rota tion Parts Inventory Rotation is defined as the annual turnover of parts held in the on -site p a rts w a reho use. No Ben c hm a rk is a va ila b le for this p a ra me ter. Defining a realistic target for this metric is highly dependent upon site logistics of the specific operation including transportation, the capacity and design of the parts warehouse, the remoteness of the site, costs associated with carrying the inventory and the specific requirements of the site in terms of a ny a va ila b ility gua ra ntee s tha t ma y b e in plac e.

2.5.6 - Em erge nc y Orders Emergency Orders quantifies the percentage of parts orders that are placed against the system on an emergency basis, i.e. “panic mode”. The p erce nta ge of Eme rg enc y Orders is a nothe r me thod of a na lyzing the extent to which the maintenance organization is behaving pro-actively a nd c on trol of the flee t. The re is no Ben c hm a rk or ta rg et for this me tric.


2.5.7 - Inve ntory (item s) This me tric q ua ntifies the num b er of individ ua l line ite ms ma inta ine d on site in the parts inventory. Because this is proportional to the size of the fleet being supported, there is no Benchmark or target for this metric. Tren d ing invento ry lev els ov er a 6 to 12 mo nth p eriod a nd relat ing the tren d to flee t p e rfo rma nc e results suc h a s MTBS, MTTR a nd % o f sc he d uled wo rk, is one wa y of d ete rmining the imp a c t of p a rts sup p ort on the ove ra ll site p erforma nc e.

2.5.8 - Invento ry (va lue) This me tric q ua ntifies the va lue of the on -site p a rts inve nto ry. Here a g a in, because this is proportional to the size of the fleet being supported, there is no Ben c hm a rk or ta rg et for this me tric . Trend ing inven to ry va lue ov er a 6 to 12 month period and relating the trend to fleet performance results suc h a s MTBS, MTTR a nd % o f sc he d uled w o rk, is o ne w a y o f d e te rmining the imp a c t of p a rts sup p ort on the o verall site p erforma nc e.

2.6 Rep a ir Ma na g em ent The follow ing Rep a ir Ma na ge me nt me tric s a re indica tors of the a d eq ua c y of the maintenance personnel, facilities, tooling and support equipment and how well those resources are organized and managed to perform efficient & effective repairs while contributing to efficiency, cost and availability objectives.

2.6.1 - M TTR (sho p servic e)

MTTR shop =

Tota l Sho p Rep a ir Dow ntime

(hours)

Numb er of Sho p Rep a irs

The a verag e d ow ntime hours (inc luding d elays) req uired to re p a irs. The re is no Ben c hm a rk or ta rg e t a sso c iat ed w ith Actual results will vary significantly based upon the nature whether it is scheduled or unscheduled and the extent to are grouped for op timum effic ienc y.

exec ute shop this m ea sure. of the repair, which repairs


2.6.2 - MTTR (field)

MTTR field

=

Tota l Field Rep a ir Dow ntime

(hours)

Numb er of Field Rep a irs

The a ve ra g e d ow ntime ho urs (inc lud ing d ela ys & resp on se time ) req uired to exec ute field rep a irs. There is no Benc hma rk a ssoc iate d with th is me a sure b ut a rea sona b le t a rge t is <1 hour. It should b e note d here tha t the field service crew is the “first line of defense” for the maintenance organization and, as such, their attention should be focused on rapid remedial actions and fast, decisive decision-making. Consequently, if the field servic e c rew d ete rmines tha t a rep a ir c a n be b ette r p erformed in the shop environment, they should release it to the shop unless those re so urc es a re una va ila b le. Thus, the t a rg et fo r MTTRfield is re lat ively low , <1 hour, and unless the machine is not mobile, higher quality and more effic ient repa irs c a n b e p erformed in the shop .

2.6.3 - M TTR (sho p servic e / without d ela ys)

MTTR shop

=

Tota l Sho p Rep a ir Dow ntime

(hours)

Numb er of Sho p Rep a irs

The a verag e d ow ntime hours (excluding d elays) req uired to exec ute shop re p a irs. The re is no Be nc hm a rk or ta rg e t a sso c iat ed w ith this m e a sure. This measure enables site management to assess the impact of delays on shop repair execution and, if delays are found to be a problem, take c orrec tive a c tions a c c ord ingly.

2.6.4 - MTTR (field / without d ela ys)

MTTR field

=

Tota l Field Rep a ir Dow ntime

(hours)

Numb er of Field Rep a irs

The a ve ra g e d ow ntime ho urs (exc lud ing d ela ys & resp on se time ) req uired to exec ute field rep a irs. There is no Benc hma rk a ssoc iate d with th is m ea sure b ut a re a so na b le ta rg et is <.5 ho ur. This m ea sure e na b les site management to assess the impact of delays on field repair execution Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

and, if delays or response time are found to be problems, take the c orresp ond ing c orrec tive a c tions.

2.6.5 - MTBS a fte r Rep a irs

Tota l Op era ting Hours to First Sto p

MTBS a fter Rep airs =

(hours)

Numb er of Rep a irs

The a ve ra g e o p erating hou rs to t he first sto p a fter ea c h rep a ir is a n ind ic a tion of rep a ir q ua lity a nd e ffec tive ne ss. There is no Benc hm a rk available for this metric. A realistic target depends upon the level of current performance, i.e. if repair quality is an issue (technicians are creating more problems than they are correcting), the target should be mu c h hig her tha n if rep a ir q ua lity is relat ive ly go od .

2.6.6 - % Red o

Redo =

Σ Rep a ir Σ

Red o Ma n-Hou rs

Rep a ir Ma n-Ho urs

X 100

(%)

Percentage Redo is a measure of repair effectiveness. It is the proportion or percentage of man-hours labor dedicated to redo (correcting repair errors) to the total number of man-hours labor dedicated to all repairs. There is no Benc hm a rk for this me tric b ut a rea sona b le ta rg et is tha t red o sho uld b e < 2% of to ta l rep a ir ma n-hours.

2.6.7 - Una va ilab ility Dela ys

Unavailability D

=

(1

Availability) X

Dela y Dow ntime Tot a l Dow ntime

X 100 (%)

Unavailability due to delay downtime quantifies the impact of delays a sso c iat ed w ith sho p a nd field rep a irs on a va ila b ility. The re is no Benchmark associated with this metric but a reasonable target is < 1%. If unavailability (downtime) due to delays is significantly higher than 1%, it ma y b e rela ted to inad eq ua te m a np ow er, exce ssive field resp onse times, poor communications to the field service crew, excessive workload, and / or a highe r than normal amo unt of unsc hed uled wo rk. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

2.6.8 - Co ntam ination Co ntrol The “ Sta r” c onta mination c ontrol ra ting sc hem e is a va lid ind ic a tion o f the organizations ability to provide and maintain a working environment and work procedures that promote and support high quality repair execution. The Be nc hm a rk is “ 5 Sta r” .

2.7 Comp onent Manage me nt 2.7.1 - Co m p one nt Life Ta rg et A c hieve d

Co mp one nt Life = Ta rg et A c hieve d

Co m p on en ts Exc ha ng ed _ > Life Ta rg et X 100

Tota l Com p one nts Exc ha ng ed

(%)

The Co mp one nt Life Ta rge t Ac hieve d q ua ntifies the p erce nta g e o f components that are reaching target life before component exchange. There is no Ben c hm a rk for this me a sure b ut a rea son a b le ta rg et is tha t 100% of c om p one nts a c hieve ta rge t life to reb uild .

2.7.2 - Co m p one nts Sc hed uled

Components = Sc hed uled

Sc hed uled Co mp onent Exc ha nge s Tota l Com p one nts Exc ha ng ed

X 100

(%)

The Co mp one nts Sc hed uled me tric q ua ntifies the p erce nta g e o f c om po nents that a re excha nge d o n a sc hed uled ba sis a s a p erc entag e of the tota l numb er of c om p one nts excha ng ed in the p eriod . There is no Benchmark for this measure but a reasonable target is that 100% of c om po nents a re e xc hang ed on a sc hed uled b asis.

2.7.3 - C om po nent Rep lac em ent Co sts

Component Replacement = Cost

Ac tua l PCR Co sts Bud g et ed / Ta rg et PCR Co sts

X 100


(%)

The Co mp one nt Rep la c em ent Co st me tric q ua ntifies the ra tio o f ac tua l c om p one nt rep la c em ent c osts to b ud g ete d (target) c osts. There is no Benchmark for this measure but a reasonable target is that actual c om p one nt rep la c em ent c osts should no t exce ed (< or = 100%) bud g et.

2.7.4 - Unava ilab ility Com po nent Rep lac em ent

Unavailability CR

=

(1

Availability) X

Co mp . Rep la c em ent DT To ta l Dow ntime

X 100 (%)

Unavailability due to component replacement downtime quantifies the imp a c t of c om p one nt rep la c em ent e vents on a va ila b ility. There is no Benchmark associated with this metric but a reasonable target is < 2%. If unavailability (downtime) due to component replacement is significantly higher than 2%, it may signify potential problems with component inventory quality / quality, component rebuild (CRC) turnaround time, and/or a higher than normal percentage of unscheduled component rep la c em ent e vents. If the c om p one nt invento ry q ua lity / q ua lity is found to be an issue, it may be due either to the fact that the maintenance organization is not doing a good job of communicating the component replacement plan the Parts Department and/or CRC or that the Parts Dep a rtme nt a nd / or CRC is not d elivering on its ob lig a tion to sup p ort the site w ith the c om p one nts they req uire.

2.7.5 - A va ilab ility Inde x a fter Com po nent Exc hang e The Ava ila b ility Ind ex d uring the first six mo nths a fter ma jor c om p o nen t exchange is a very good indication of the quality of work being pe rforme d d uring c omp onent excha nge . This c a n be c alcula ted by monitoring Availability Index for the individual machine for the period. There is no Benc hma rk for this me tric b ut a va lid ta rg et is tha t the Availability Index during the first six months after major component exc ha ng e is > 88%.

2.7.6 - PCR Co m p lianc e

PCR’ s Pla nne d , Sc he d uled & Exec ute d PCR = Compliance To ta l PCR’ s Pla nned & Sc he d uled

X 100


(%)

PCR Compliance defines the number of components that are replaced on a planned and scheduled basis as a percentage of the total number of c om p one nts p la nned a nd sc hed uled . There is no Benc hma rk for this measure but a reasonable target is that 100% of components that are planned and scheduled for replacement are indeed executed per the plan.

2.7.7 - Bac klog s Exec uted during Com po nent Exc hang e

Backlogs B’ lo g s Exec ute d d uring C om p o ne nt Exc hg Executed = Tota l Ba c klo g s Pend ing w ith PCR

X 100

(%)

Backlogs executed during component exchange is a good indication of how we ll the orga niza tion is using the “ windo w of o p p ortunity” p resente d by c omp onent exc hange to bring the eq uipm ent up to a standa rd that w ill pe rm it it to p e rfo rm reliab ly. The re is no Ben c hm a rk for this m e a sure b ut the target should be that 100% of items on the pending Backlog list are executed d uring the co mp onent rep lac eme nt ac tivity.

2.8 Hum a n Resourc es a nd Training 2.8.1HR - Una va ilab ility Personne l Dela ys

Unavailability PD

=

(1

Availability) X

Personnel Delay Downtime To ta l Dow ntime

X 100 (%)

Unavailability due to personnel delay downtime quantifies the impact of p erson nel d elays (wa iting on ma np ow er) on a va ila b ility. There is no Benc hm a rk assoc iate d with this me tric b ut a rea sona b le ta rg et is < .5%. If unavailability (downtime) due to personnel delays is significantly higher than .5%, it may signify potential problems with manpower staffing levels, higher than normal workload (an abundance of Backlogged repairs), workforce inefficiency, and/or a higher than normal amount of unscheduled work, i.e. work that enjoys the benefits of planning & scheduling are typically executed far more efficiently than work that is Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

performed in an environment of constantly changing priorities such as the one c rea ted when a high p erc entag e of the wo rkloa d is unsc hed uled .

2.8.2HR - Effec tive M a n-Hours

Effec tive M a n Hours

=

Σ La b o r Ho urs (han d s o n) p resen t in the m on th [ h ]

The numb er of ha nd s-on la b or hours a va ila b le p er month is a me tric tha t help s und ersta nd / me a sure th e m a np ow er ava ila b le o n site . This me tric is use a s a refe ren c e for othe r me tric s.

2.8.3HR - Personne l Ha nds-On to Ma c hine Ra tio

HO Personnel – Ma c hines Ra tio

Tota l Ha nd s On Tec hnic ia ns =

(-)

Numb er of Ma c hines

This ra tio g ive s us a q uick rela tionship b et we en the w ork forc e on-site a nd the num b er of m a c hines. This me tric c a n be used a s a q uic k referenc e and must be combined with other metrics to better understand the p erso nne l situa tion o n-site.

2.8.4HR - Direc t Maintena nc e Personne l

(-)

Direc t Ma intenanc e = Qua ntity o f Ha nd s-On Tec hnic ia ns Personnel This me tric d oc um ent s the num b er of ha nd s-on te c hnicia ns. The t ota l numb er of tec hnic ia ns on-site c a n be ea sily c om p a red with other metric s a nd ot her sites with simila r op erationa l c ha ra c te ristics.

2.8.5. HR - Personn el Tota l to M a c hine Ra tio

Total Personnel – M a c hines Ra tio

To ta l Pe rson nel =

Num be r of M ac hines


[-]

This m et ric e sta b lishe s the rela tionship b et w ee n the to ta l p e rson ne l o n-site a nd the num b er of ma c hines. Tota l p ersonne l inc lud es a ll p eo p le o n site, technicians, planners, supervisors, secretaries, administration personnel, etc.

2.8.6HR - Tota l Ma intenanc e Personne l To ta l Ma inte na nc e Personnel

=

Tot al Qua ntity of M a inten anc e Personnel

[-]

This me tric a sks for the to ta l ma inte na nc e p erson ne l on-site . How ma ny p eo p le we ha ve on-site to sup p ort the flee t o f eq uip me nt. This me tric in c o njunc tion w ith me tric s in 2.8.3, 2.8.4 & 2.8.5 g ives us a d ime nsion o f the maintenance crew on-site and can be compared with other similar op erations to sup p ort the und ersta nd ing of p ersonne l effic ienc y. All the above metrics are very straightforward and can be easily ob ta ined . The d oc ume nta tion of these fig ures link with the c ha ra c teristic s of the fleet and the on-site operation will give us a quick reference for future studies.

2.8.7HR - Ove rtime to to tal Effec tive Hours Ra tio

Ov e rtim e ho urs in the Pe rio d Ov e rtim e Ra tio

=

X 100 (%)

To tal e ffec tive ho urs in the Pe riod The ov ertime ra tio m ea sures the a mo unt o f extra ho urs req uired to sa tisfy the maintenance operation as a percentage of the total effective hours ava ila ble in the p eriod . Overtime ma y be an indica tion o f lab or need s or extra maintenance demand in a particular period of time. Constant overtime levels should be an indicator to consider to review and possibly to im p rov e t he on-site o rg a niza tion struc ture.


2.8.8HR – Labor Hours Covered by Workorders

La b o r Ho urs in Wo rk Order

Total Labor Hours in Workorders

X 100 (%)

=

To tal e ffe c tive ho urs in the Pe riod

The la b or hours c a p tured in the wo rkord ers a re c ha rg ed to sp ec ific machines whenever a technician is assigned and working on that p a rticula r ma c hine. This ra tio is a n indic a tion of te c hnicia n idle time. Id le time inc lud es b rea ks for lunc h a nd shift c ha ng es. The ide a l situa tion w ou ld b e to ha ve a ra tio ve ry c lose to 100%, ind ic a ting a full usa ge of the la b or available. Be careful in those cases in which the on-site operation “justifies” technician labor hours by charging 100% of their hours to the ma c hines b eing servic ed .

2.8.1T – Training Hours / La b or Hours

Training Ho urs Ratio

To ta l Training Ho urs in th e Pe rio d =

X 100 (%)

To tal La b o r Ho urs av ailable in the Period

This m et ric, use d b y the Inte rna tiona l La b o r Org a niza tion, g ives us the relationship b etw ee n the hours invested in training a nd the tota l ava ila b le la b or ho urs in the p eriod . This orga niza tion c on side rs tha t a minimu m of 1% of every 100 lab or ma n-hours be d ed ic ate d to tra ining to be ac c ep tab le. We d on ’ t ha ve results for our mining p rojec ts b ut w e b eliev e tha t if it we re known it would b e m uc h hig her tha n the OIT ta rg et.

2.8.2T – Training Hours b y System s (vs. Prob lem s)

Training Hours Ratio by Systems

=

Tra ining Hours on “ Syste m A”

X 100

Tota l Training ho urs in the Period

This m et ric g ives us the o p p o rtun ity t o a sse ss if the training effo rts a re addressing the actual problems that the maintenance organization is fa c ing. The relat ion ship b et we en the Ra nking of Prob lem s (Top 5 or 10) obtained from the machine repair history and the hours trained in that p a rticula r sub jec t or ma c hine a rea must be ve ry strong . The tra ining nee d s d ete c tion (TND) proc ess sho uld c onside r this ra nking of p rob lem s Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

(%)

and schedule training activities to support the organization to be better p rep a red to solve tho se p rob lem s.

2.8.3T – Tra ining Prog ra m Co mp lianc e

Training hours c om plete d

Training Prog ram Compliance

=

X 100 (%)

Total training hours sc he duled

The training c om p lia nc e m etric gives us a n ind ic a tion o f how we ll the o nsite op e ra tion is fo llo w ing the tra ining a c tivities sc he d uled . It is a lso a n ind ic a tion o f the imp ortanc e tha t is g iven to the training a c tivities. Are w e c om plying with what we sc hed uled ?

2.8.4T – Co m p ete nc ies Req uired

Competencies Required

Com pe tenc ies de fined in ea c h role & re sp o nsibilitie s X Num b e r o f Pe o ple in the c o rresp o nding Roles

=

To o b ta in this me tric , to ta l num b er of c om p et enc ies req uired , the o n-site organization must have a clear definition of the roles and responsibilities for each of the positions in the organization and the number of people a ssigne d to ea c h o f tho se p a rtic ular ro les. This m e tric g ives us a n indica tion a s to how ma ny co mp etenc ies we need in orde r to d o the job effic ienc y a nd effec tively.

2.8.5T - C om p ete nc ies Sa tisfied

% of Competencies Satisfied

Pe rso nnel C om p e te nc ies C e rtifie d =

X 100 (%)

Co m p e tenc ies Re q uire d

The p erc ent a g e o f Co mp et enc ies sa tisfied will g ive us a n und ersta nd ing of how well the organization is prepared to perform the maintenance ta sks. This me tric c om p a res the num b er of c om p ete nc ies tha t have b ee n certified in a formal certification process against the total competencies required. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

2.8.6T - Personnel Skills Inventory Com p letio n

Skills / C om pe te nc ies in Inve nto ry

Personnel Skills Inventory

=

X 100 (%)

To tal Pe rsonnel Skills

The Pe rso nne l Skills Inve nto ry is p ro b a b ly th e fou nd a tion for a ny t ra ining d e te c tion ne e d s o r g a p a na lysis stud y. The p e rso nne l skills m ust b e surveyed and kept in a n upd a ted Data Base.

2.9 Continuous Im p rovem ent 2.9.1 - Top Prob lem s with C.I. Proje c ts

C I wo rking o n Ta rg e t (CIOT)

=

% of Co ntinuous Im p rov em e nt Pro jec ts o n Pro b lem Ra nking To p 1-2

_________ %

To p 3-5

_________ %

To p 6-10 _________ % The Ra nking o f Pro b le m s is b a sed o n the Ma c hine Rep a ir Histo ry a nd rep resent s tho se p rob lem s or area s tha t mo st a ffec t the final pe rforma nc e of th e flee t. The % of C I p rojec ts d ed ic a te d to solve th ose p rob lem s is a n indica tion of effec tive c om munic ations a nd c lea r ma nag eme nt to direc t the o rg a niza tion e fforts whe re the y are mo st nee d ed .

2.9.2 - Ac tive C.I. Projec ts

Active C I Pro je c ts

=

Num b e r o f C I p ro jec ts in p roc ess [ - ]


This me tric , the num b er of C I p rojec ts tha t a re in p roc ess or a c tive, gives us an unde rstand ing o f the ma gnitude of the C I effort a nd how a c tive the o rg a niza tion is in t he sea rc h fo r so lutio ns o r op timiza tions.

2.9.3 - C.I. Projec ts Closed d uring Period

C I Projec ts Closed

=

Num b er o f C I p ro jec ts c lose d in the Pe riod [ - ]

This me tric is a n indic a tion o f the reg ularity o f the CI proc ess in plac e, tha t is how ma ny CI p rojec ts a re w e c losing in a p eriod of time, or how m a ny solution proposals are we delivering to the organization. We recommend m ea suring this re sult o n a m o nth ly ba sis.

2.9.4 - A ve ra g e C.I. Projec t Exe c ution Time

CI Projec t Execution Tim e

=

Averag e d a ys to c om p lete CI p rojec ts

[d a ys]

This me tric in c on junc tion w ith 2.9.3 ab o ve g ive s us a me a sure of t he CI p roc ess effic ienc y. Ave ra ge d a ys to c lose the C I Projec ts c oup led with the regularity of delivery measured by CI Projects Closed in the period provides us a means of telling if the CI pipeline has the ability to deliver results in a rea son a b ly c on sta nt time fra me .

2.9.5 - C.I. Projec ts Op ened d uring Period

C I Projec ts Opened

=

Num b e r o f C I Pro jec ts g e ne ra ted in the Period [ - ]

The num b er of CI p rojec ts op ene d or gene ra ted in the p eriod is a n ind ica tion o f ma turity o f the CI proc ess. This m et ric m ea sures if the maintenance organization is continually looking for solutions and better ways to do things. It is also is an indication of the effectiveness of the Prob lem / Op po rtunity de tec tion p roc ess a nd ma nag em ent. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

2.9.6 - C.I. Projec t Delivered Va lue

C I Delive re d Value

=

Prov e n Bene fits d ue to CI solutio ns Implemented

This me tric me a sures the effe c tivene ss of the CI proc ess in p la c e throug h the p rove n va lue d elivered to the op eration. This va lue c a n b e a recorded direct savings, an increase in productivity, an improvement in availability translated into more equipment hours to produce and / or ma intenanc e la bo r red uc tion, etc. Ac c ura te rec ord s are a m ust in orde r to b e a b le to c a lc ula te a nd va lid a te the results of this me tric .


[$]

Glossa ry of Terms Performance Metric: A term used to describe the outcome of any p roc ess used to c ollec t, ana lyze, interpret a nd p resent q ua ntita tive d a ta . A m ea surem ent p arame ter that ena b les p erforma nc e a ga inst som e pred efined Ta rge t or Benc hma rk to b e mo nitored. A me a surem ent used to gauge performance of a function, operation or business relative to past results or o th e rs.

Key Performance Indicator: Also known as KPI; a top level Performa nc e Me tric. The c ollec tion of KPI's used to d esc ribe p erforma nc e of a particular project may vary from site to site, by product, application and even one's perspective, i.e. dealer & customer, Operations & Ma inten a nc e De p ts., Pro jec t & C o ntra c t Co ntrols Dep t. NOTE: All KPI's a re Performa nc e Me trics b ut Performa nc e Me trics a re no t a lwa ys KPI's.

Ta rget: A d esired go a l; a stand ard b y whic h a Performa nc e Me tric c an b e me a sured or jud g ed . The Ta rg et for a p a rticula r Performa nc e Me tric can be somewhat arbitrary and will likely vary by product, application or sp ec ific site . The Ta rg et is freq ue ntly de termined b y c usto me r nee d s, his expectations and / or contractual commitments, and manufacturers’ specifications.

Benchmark (noun): A world-class performance standard relative to a sp ec ific Performa nc e Me tric; rep resent s a nd q ua ntifies "b est p ra c tice " of an operation or of specific functions within that operation according to a sp ec ified Performa nc e Me tric . A Benc hma rk ma y vary by prod uc t but, b y c o ntra st, is m uc h le ss a rb itra ry tha n a Ta rg et . A Be nc hm a rk is d e te rm ined by and represents actual, documented, sustainable performance over time relative to som e Performa nc e Me tric .

Shutdow n / Stopp ag e: An event that takes a machine out of service. Shutd ow ns ma y be sc hed uled or unsc hed uled a nd inc lud e a ll type s of maintenance and repair activities except daily lubes, refueling and insp ec tion s exec ute d during lub e or refue ling ac tivities. Op erationa l stoppages, e.g. shift changes, lunch breaks, etc., are not included as shutd ow ns. “ Group ed ” rep a irs c ou nt a s a sing le shutd ow n. Shutd ow n count is independent of event duration or complexity, i.e. a five-minute event c ounts the sam e a s a 100-hour event a nd a hea d light replac em ent c ounts the sa me a s a c a ta strop hic ma jor c om p one nt failure. Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Failure: The inab ility of a d ev ice o r syste m t o p e rfo rm in its re q uired function … the cessation of proper functioning or performance, for exam p le a p ow er fa ilure. Fa ilures ma y b e the result o f tec hnic a l prod uc t issues, i.e. eq uipm ent un relia b ility, or due to m a inte na nc e / rep a ir neg lec t, i.e. equipment management ineffectiveness in the area of problem avoidance.

Tota l Ca lend a r Hours: Tota l time in the p eriod to b e a na lyzed , e.g . 8760 ho urs / ye a r, 720 hours / 30 d a y mo nth , 168 hours / w e e k, e tc .

Scheduled Hours: Time tha t a ma c hine is sc hed uled for op erations. Typic a lly de termined b y the m ine Pla nning a nd Op erations Dep a rtme nts in c onjunction w ith the ir overa ll prod uc tion ta rg ets.

Unscheduled Hours: Hours outside the plan; lost time that result from accidents, strikes, weather, acts of God, any holidays that are observed, etc. (typically defined by the customer or contained in the Customer Sup p ort Ag ree me nt or MARC).

Available Hours: Time tha t a m a c hine is c a p a b le o f functioning in the intende d op eration.

Operating Hours: Time tha t a ma c hine is a c tua lly op erating in the intende d function.

Stand-by Hours: Time tha t a ma c hine is a va ila b le fo r op eration b ut is not b eing used , e.g. no op erato r a va ila b le, "ove r-truc ked", etc . know n a s "Rea d y line " hours.

Also

Production Delay Hours: Time tha t a ma c hine is op erational b ut is waiting with the engine running due to blasting, loader wait time, etc. Production delay hours are frequently not accounted for separately and a re inc lud ed in the op erating hours ta b ula tion. One the other ha nd , some dispatch systems do track production delay hours in an effort to minimize a nd m a na g e the m. In either c a se, lost hours tha t result from production delays should be reconciled and not counted against ma c hine a va ila b ility.

Operational Delay Hours: Time tha t a ma c hine is a va ila b le fo r operation but is not being used due to shift changes, lunch breaks, me et ing s, p ra yers, et c . Just a s w a s the c a se fo r p rod uc tion d elay ho urs, Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

lost hours that result from operational delays should be reconciled and never c ounted ag ainst ma c hine a vailab ility. On the other hand, po lic y at many mines ignores operational delay hours altogether and therefore, does not credit operational delay hours as either scheduled or available hours.

Downtime Hours: Time tha t a ma c hine is not a va ila b le for op eration; out of service for all forms of maintenance, repairs and modifications. Inc lud es insp ec tion a nd d ia g nostic time a s we ll as a ny d elay o r wa it time for manpower, bay space, parts, tooling, literature, repair support eq uipm ent, d ec ision ma king, etc. Ma y be sc hed uled or unsc hed uled .

Repair Delay Hours: Time tha t m a c hine is wa iting for rep a irs d ue t o una va ila b ility of lab or, p a rts, fa c ilities, eq uipm ent o r to oling . Typ ica lly not well documented in most machine downtime histories but is nonetheless inc lude d , yet unrec og nized , as p art of the ma c hine d ow ntime rec ord .


Ge neric Pa reto Referenc e for La rg e O ff Highwa y Truc ks The m ost suc c essful mining sup p ort op erations a re th ose th a t ha ve a c lea r und ersta nd ing of the prob lem s a nd issues the y a re fac ing . The id entific ation and qua ntific ation of p rob lems by c om p onent (e.g. engine, transmission, …), system (e.g. hydraulics, electrical, …) or even process (e.g. PM) facilitates an understanding of the influence each is having on the final outcome enabling management to focus its attention and resources on the most critical issues, i.e. those that will derive the ma ximum be nefit. Unfortuna tely there a re no Benc hma rks tha t a re a p p lic a b le to this kind of me a surem ent. How eve r a c ollec tion o f guidelines for la rg e Off Highw a y Truc ks in the 785 – 793 size c lass is a va ila b le to ev a lua te a c tua l site p e rfo rm a nc e in te rm s o f MTBS, MTTR a nd imp a c t o n Av a ila b ility. While no t actually benchmarks, the information contained in this reference defines what we believe to be a reasonable level of acceptability for frequency o f d o w ntime e ve nts (MTBS), d urat io n o f d o w ntime e ve nts (MTTR) an d imp a c t on Availa bility for ea c h of the ma jor area s on the ma c hine. The d a ta is rep resent a tive o f a site o p erating a t a n Ava ila b ility Ind ex of a p p roxima te ly 90% a nd is, of c ou rse, g en eric sinc e a c tua l results a c hieved at any given mine are site-specific because results of this kind are a function of not only application severity but also of the operating environment, the maintenance the equipment receives and product d esig n shortco mings tha t ma y b e p ec ulia r to spe c ific ma c hines either b y mo d el or within a g ive n ra ng e o f seria l num b ers. The ta b le on the follow ing pa ge p rovide s a ba seline tha t c a n b e used a s a reference until individual site experience and history can be do c umented . Using it to ge nerate and p erform a top problems d istrib ution a na lysis ena b les Projec t M a na g em ent to id entify a nd p rioritize critical issues affecting success of the project for investigation and resolution. It is imp ortant to note tha t d ue to the fac t that m ost ma c hine w orkorde r history d ata b ases to no t d o a very go od job of d oc umenting de la ys, the d ura tion of m a c hine do w ntime e ven ts (MTTR) inc lud es a ny d elay d ow ntime a ssoc ia ted with the pa rtic ula r eve nt. As suc h, d elays a re no t listed sep a ra tely in the ta b le. A b ette r wa y of d oing b usiness wo uld b e to break out actual repair downtime from any delays associated with the Caterpillar Confidential Yellow Copyright 2007 Caterpillar. All Rights Reserved. Proprietary Material of Caterpillar Inc. The material contained herein is confidential and is not to be copied or distributed without written permission by owner

Metrics (KPI's) to Assess Process Performance

Recommend Documents