WORK MEASUREMENT Rodger Koppa, P.E., Ph.D. Industrial and Systems Engineering
Uses of Work Measurement Compare efficiency of alternative methods Balance work among team members Optimize number of machines per operator (e.g., in a work cell) Establish basis for
1.
2. 3.
4. • • • •
Production planning and control Layout Process planning JIT
More Uses 1. 2.
3.
Pricing estimation Standards for labor performance and machine use for 1-5 and for incentives Information for labor cost control – enable standard costs to be fixed and maintained
Basic Six Pack
SELECT what to study RECORD that activity or operation EXAMINE the recorded data and modify procedures/task allocation/layout using Work Study to get the best method MEASURE quantity of work with respect to time with the best method or ESTIMATE using predetermined time data COMPILE a standard time for activity, including allowances DEFINE activity and standard time
Work Measurement Techniques 1. 2. 3. 4.
5.
Work Sampling Structured Estimation Time Study Predetermined Motion Time Standards (PMTS) Standard Data Systems (SDS)
Technique 1: Work Sampling
Definition—Work Sampling
Work sampling is a method of finding the percentage occurrence of a given activity by statistical sampling and random observations Also known as:
Activity sampling Ratio-delay study Random observation Snap-reading method Observation ratio study
When to Use Work Sampling
Lots of time available (weeks) More than one worker and/or machine Long cycle times Non-repetitive work cycles (but must be distinct number of categories)
Statistical Approach
Best Case: observe activity all the time Worst Case: observe activity once and jump to conclusions based on that one experience Realistic Case: observe often enough to draw some conclusions with a given level of confidence (80%, 90%, 95%, 99%.....) How Often?
Setting How Often
How much confidence you need determines how many times you must observe and go through the six-pack Sometimes the target operation will be going on, sometimes not Spend a period of hours or visit 5 times or more to estimate % time activity happens
Sample Size Estimation Where σp = SE of proportion p = % idle time q = % working time n = Number of observations needed (sample size)
σp =
pq n
Sample Size Estimation (Cont’d)
Suppose 95% confidence level 10% margin of error Then 1.96 x (σp) = 10; so σp ≈ 5 Preliminary study says machines idle 25% Substitute these values of σp, p,q in binominal sample equation and solve for n n = 75 observations with margin of +/- 10% (if +/- 5%, you need 300 observations!)
Randomized Observations (Play the game right)
Randomize with respect to time Divide shift into time slots Use random number table (or Excel) and assign random numbers to each time period, re-order time slots 1-n Observe in that order (will take quite a few days or maybe weeks)
For Each Observation
Record what is happening to level of detail required: Machine working?
Cutting Boring Grinding
Machine Idle?
Maintenance Waiting for materials Worker in restroom
Rated Work Sampling
Equal-appearing rating (≤ 10 intervals) can be applied to pace of operation if machine not idle Can set fixed intervals (no random sampling) and rate pace at those times Interval must be short with respect to total cycle time Can be used with more than one worker and/or machines
Work Sampling: So What?
Simple technique usable in plants, service operations, offices Low cost Avoids controversial aspects of time study Management gets good idea of where inefficiencies may lie Can trigger method studies, travel studies
Work Measurement Techniques • • • •
•
Work Sampling Structured Estimation Time Study Predetermined Motion Time Standards (PMTS) Standard Data Systems (SDS)
Str uct ur ed Est imation Most widely used and oldest approach Based on past experience with same/similar operations Very cheap to apply, but you get what you pay for Much used for jobs/operations not yet implemented
An al ytical A ppr oac h t o Str uct ur ed Est imation Estimator should have experience in type of job being considered and in work study techniques (tall order) Alternatively the estimator derives times/rates of production from experienced workers by debriefing
An al ytical A ppr oac h Break job into elements than can estimated for time Apply available estimates (experience, Dodge Manual, other labor statistics) Time similar task elements or use mockup/simulated workplaces SWAG
Work Measurement Techniques • • • •
•
Work Sampling Structured Estimation Time Study Predetermined Motion Time Standards (PMTS) Standard Data Systems (SDS)
What is a “Time Study?” “Time study is a work measurement technique for recording the times of performing a certain specific job or its elements carried out under specified conditions, and for analyzing the data so as to obtain the time necessary for an operator to carry it out at a defined rate of performance” Introduction to Work Study 4th Ed G. Kanawaty (Ed.) International Labor Office, Geneva Switzerland 1992
Basic Time Study Approach Except for a Q&D study, don’t bother with a stop
watch
Heisenberg principle Requires practice and multiple runs
Acquire and use a video camera Digital video camera best and getting cheaper every year Many digital (still) cameras have “movie” mode that suffices for short cycles (5 min or less) “Movie Maker” software to analyze
Data Collection Can make up your own or use a spreadsheet Basic sheet layout sketched in next slide Summary sheet also can be hand-hacked or
produced on a spreadsheet (much prefer using a spreadsheet!)
When to Do a Time Study Don’t bother with time study until Method
Study is done and as much inefficiency is wrung out as possible Must optimize method before setting a standard! Aim for what a “qualified” worker can do without “undue” fatigue (engineering judgment)
Caution Don’t try to do covert observations either with
stopwatch or with camera Invasion of personal privacy May or may not meet criterion of “motivated” worker Duck image of “efficiency expert”
8 Basic Steps Identify background, operator, environmental factors that pertain to work 2. Do a Task Description down to “element” level of detail 3. Do a Methods Study and optimize the task 4. Time each task element from video data or by direct observation with a stopwatch 1.
8 Basic Steps (Cont’d) • • • •
Assess observed pace vs expected pace (IF past experience/standard times exist) Extend observed times to Basic Times Determine Allowances Establish Standard Times
1. Background Information Retrieval/catalog information Study number Analyst and approvals (if needed) Date, time, location Product/part identification Process, method, plant, machine identification Worker’s identification or designation (do you really
need this?) Study duration Working conditions (temperature, humidity, lighting, noise)
2/3. Task Description & Methods Study Element: lowest component of a specified job
being timed, selected for convenience of observation, measurement, and analysis Work Cycle: sequence of elements required to perform a job or yield a unit of production
2/3. Task Description & Methods Study (Cont’d) Observe task being done several times Debrief worker based on observations to identify
subtasks and/or study videos Perform Methods Study Optimize task based on results of Methods Study Train worker(s) to perform optimized task Repeat task cycle until performance becomes asymptotic in skill and time Complete task description down to element level
2/3. Task Description & Methods Study (Cont’d) Types of Elements: Repetitive – comes up each cycle Irregular – comes up every now and then Constant – takes same time every time Variable – basic time varies with respect to some characteristics of product Manual – done by worker Machine – automatic once started, can only be terminated by operator Governing – Takes more time than any other element within work cycle performed concurrently
2/3. Task Description & Methods Study (Cont’d) Picking elements to study: Define break points—distinct segments Elements should be 2-3 sec or longer For timing purposes: Separate manual and machine elements Separate constant from variable elements Separate repetitive from irregular elements
4. Perform Time Study How many repetitions to time or video? 2 Approaches: 3. Statistical estimation of sample size 4. Industrial practice (preferred) Industrial Practice (source: General Electric) The shorter the cycle time, the larger the sample size (Number of reps) Vary from 0.10 min = 200 reps to 40 or more min = 3 reps
Timing Each Element Video Recording: Record as many cycle reps as needed for sample Use time scale in software (Windows Movie Maker) Stopwatch (ugh): Cumulative timing: note start time of cycle and then note time at end of each element Flyback timing: watch reset (but not stopped) at end of each element and time noted Digital watch MUCH easier than analog Will have to go through cycle as many times as needed for sample size and repeat for each concurrent element
Time Study Rating Very subjective and full of pitfalls Helps if observation is as unobtrusive as
possible without being covert Helps if enough reps done so worker forgets you are observing him AND if the worker is “qualified”
Qualified? A qualified worker is one who has acquired the skills, knowledge and other (?) attributes to carry out the work in hand to satisfactory standards of quality, quantity, and safety Bootstrap? You bet!
Experience to Become Qualified May take only 1 shift May require 10,000 cycles for a complex
operation Motivation of worker key element Adversarial image of industrial engineer an impediment: How do you make a win-win situation? How much do you know about the operation?
Fair Time Come up with factor to multiply observed
mean time to estimate “fair time” “Standard performance” is what can be expected of an “average” worker over the shift period, assuming: Knowledge of the job (qualified) Motivation to do work in professional manner
Rating of such performance = 100
Rules of Thumb Standard rating (100) is represented by the
speed of motion of the limbs of a person walking at 4 MPH (6.4 kph) Brisk, purposeful walk Standard rating, arm motion: Dealing a pack of cards in 0.375 min
Rating Actual Performance 100 scale actually 10 scale Ordinal scale of measurement Rater must be experienced and very familiar
with operation
Intervening Variables Variation in quality of materials or
subassemblies Changes in efficiency of tools and machines Method changes Worker attention and motivation Working environment Speed/accuracy tradeoff
0 – 100 Rating Rating Performance Walking 50 Very slow, clumsy 2 mph fumbling, unmotivated 75 Steady, deliberate, unhurried 100 Brisk, businesslike piece work pace 125 Very fast, assured, dexterous coordinated 150 Exceptionally fast, intense concentration, hard to keep up pace
3 4 5 6
Deriving Standard Times for Elements Convert observed times to Basic Times
BT = Observed Time x Observed Rating Standard Rating (100) Derive “representative” BT
Mean (but consider outliers) Median or mode
Variable Elements Compute SD as well as Mean Continue to observe until SD stabilizes Use mean as best estimate of time
Work Measurement Techniques • • • •
•
Work Sampling Structured Estimation Time Study Predetermined Motion Time Standards (PMTS) Standard Data Systems (SDS)
Predetermined Motion Time Standards “ An organized body of information, procedures, techniques, and motion times employed in the study and evaluation of manual work elements. The system is expressed in terms of the motions used, their general and specific nature, the conditions under which they occur, and their previously determined performance times” --ANSI Standard Z94.11- 1989
History
1915: PMTS presaged by Gilbreth’s “Therbligs” 1927: A.B. Segur—”The time required for all experts to perform the fundamental motions (of work) is a constant” 1934: J.H. Quick – Work Factor System 1946: H.B. Maynard and others develop MTM (Method-Time Measurement) 1972: K.B. Zandlin develops Maynard Operations Sequence Technique (MOST) from MTM approach
Basic PMTS Approach 1. 2. 3.
4. 5.
Do task description in terms of basic motions in a defined workspace just as in Time Study Perform work studies to improve task (unless task is in planning stage) Retrieve from data base the basic times associated with each motion as modified by task conditions and work variables Sum all of these times to form the basic time for the task (ratings automatically included) Apply allowances (to be discussed) to arrive at standard time for the task
PMTS Levels Level 1: Very elemental motions like Therbligs, suitable for short cycles Level 2: Some motions combined, e.g., reach and grasp become get Level 3: May only be 3 or 4 elements, e.g., handle, transport, step/foot motions, bend/rise
A Few PMTS Variants
Work-Factor (1930’s) WF Methods-Time Measurement (MTM-1, MTM-2, MTM-3, others) Basic Motion Time Study (BMTS) based on MTM-1 and Therbligs Master Standard Data (MSD) based on MTM-1 Maynard Operation Sequence Technique (MOST) emphasizes production and material handling at Level 1 and 2 Modular Arrangement of Predetermined Time Standards (MODAPTS) based on MTM’s and MSD; finger motion is basic element
Basic MTM Postulates: A given motion has a fundamental time to accomplish that motion Each motion is independent of any other with respect to time Simple chaining can provide a consistent and accurate prediction of the basic time of a given operation
Time Units Vary from system to system Most common: Time Measurement Unit (tmu) = 0.00001 hour 0.0006 min 0.036 sec
MTM-2 Most widely used MTM Technique 15 basic motions Modify times by distance moved and weight Not as complicated as it looks but requires practice
Motions in MTM-2 Refer to handout, MTM-2 Basics GET– reach, grab, release GA –No grasp needed GB – grasp, close hand GC – thumb, finger only GW – grasp and move mass
Motions in MTM-2 PUT – Move object from a to b PA – ballistic motion PB – Controlled motion PC – Complex motion, obvious correcting PW –Moving significant mass
Motions in MTM-2 REGRASP (P) –change mode of grasp APPLY PRESSURE (A) –Isometric application of force EYE ACTION (E) – Examine object, search FOOT MOTION (F) – Shift foot position less than 30 cm STEP (S) –Displace trunk, leg motion more than 30 cm BEND & ARISE (B) – Lowering trunk, reach at/below knees, return CRANK (C) – Circular motion of handle
MTM -3 Bare-bones “quick and dirty” analysis Only 4 work elements:
Handle
(HA and HB) Transport (TA and TB) Step and foot motions (SF) Bend and rise (B)
A and B for H and T refer to “light” and “heavy” loads
MOST Maynard Operation Sequence Technique Basic MOST
Motion
aggregates concerned with material handling and moving objects General move-object is freely moved through space Controlled move-object is slid or manipulated Tool use-hand tool operations
Basic MOST General Move A – Action Distance: hand translation loaded or unloaded, feet translation B –Body motion sit/stand G –Gain control (same as grasp) P –Placement: position, orient, lay aside object
Basic MOST Controlled Move Object being moved is constrained (e.g., a contrtol lever) A, B, G as in General Move M –Move, controlled: object is moved over a certain path X –Process time associated with machine I –Align: motions at end of movement to ready object for next operation
Basic MOST Tool Use A,B,G,P as in General and Controlled Move Additional subtype is the specific use of the tool: Fasten,
Loosen
Cut Surface
treat Measure Record Think
Other MOSTs MaxiMOST – for long, complex cycles, e.g., heavy assembly, machine setup MiniMOST – Very short cycle, highly repetitive work (1.6 min or less) Clerical MOST – For office operations
MOST for Windows
Basic, Maxi, or MiniMOST 2 Modules: Quick
MOST – Select work methods given application, canned procedures and times edited by user as needed. Similar to using standard data systems Direct MOST – User provides data on industry and work situation, program generates possible work method. Method refined by user (intelligent system)
Work Measurement Techniques • • • •
•
Work Sampling Structured Estimation Time Study Predetermined Motion Time Standards (PMTS) Standard Data Systems (SDS)
Standard Data Systems Industry
or company specific data base Common elements from different tasks are grouped and summarized Industry sources may have data that can be applied to make a good-enough estimate of cycle/operation times MOST for Windows is such a data base Semi-quantified experience
Data Base Development
Decide what to cover, what to omit (otherwise data base will be too large) Break jobs into elements that are common enough to group together Derive times from standard data handbooks, previous time or PMTS studies on similar operations, or very limited observations (video tape if possible) of representative elements in each group
May need to use mockups or “air guitar” to get a feel for the times Estimates from workers familiar with the tasks can be very helpful, but NOT sufficient
Elements to Include Setup
and Production Constant and variable Worker-paced and machine-paced Regular and irregular
Setup and Production Elements How
long does it take to set up for any batch and changeover from one to another? Setup elements occur once per batch Production elements occur once per unit output Total time to produce a batch is the sum of setup time plus the product of number of units in batch times batch time
Constant and Variable Elements Constant:
replace bit in drill press; push 4 buttons to initiate process Variable: material handling; place work piece in lathe and secure; unload parts from container
Worker-paced and Machinepaced Elements Worker-paced:
rebuild injector body; lay up ceramic for firing Machine-paced: feed and speed, number of operations and how controlled
Regular and Irregular Elements Regular:
happen every cycle Irregular: not every cycle, so time must be prorated based on expected frequency of occurrence
From Basic Time to Standard Time • Once you have basic times from any
of the 5 approaches for each element of interest you’re not finished yet! • To derive a standard time for planning, costing, and payroll, allowances must be made • Then basic time is adjusted by allowances to arrive at standard time
Allowances Individual (e.g., disabilities) 2. Work Factors 3. Environmental Factors Cycle time is adjusted (longer) by allowances to obtain Standard Time for cycle or activity Rest periods may also be built into a shift or the time prorated over cycles as allowances 1.
PFD Allowance Personal, fatigue, and delay allowance Personal: 5% or more if stressful environment Fatigue: 5% up to 20% or more for heavy labor (negotiated or use formulas) Delay: (unintentional, caused by breakdown or inefficiencies) depends on company experience
Other Allowances Contingency: Unscheduled maintenance, breakdowns, out of tolerance products, should NOT exceed 5% Policy: machine part of cycle, training, OJT Special: Industry specific such as exposure to toxic materials, radioactivity
Machine Allowance
A “Policy allowance” negotiated between Tstd = Tnt (1 + Apfd ) + Tm (1 + Am ) management and labor Where T = worker time during cycle Pertains to wage T = machine time during cycle A = PFD allowance incentive program A = Machine allowance Machine time (automatic) “significant” part of cycle time nt
m
pfd
m
Allowance Example Assembly of a framus Cycle basic time is 10 min PFD allowance is 10% (no definite breaks, worker is free to go to restroom, water cooler) Contingency allowance is 10% (junky equipment up the line) Machine allowance is 0 (hand and pneumatic tools only in this operation) Training allowance is 5% (high turnover) Standard time = 10 x 1.25 = 12.5 min
Worker-Machine Standard Times Some terminology: Unrestricted work: hand/power tool work where output affected by factors under worker control during cycle. Personmachine relationship Restricted work: Machine does some to most of work done, worker initiates, terminates, but cannot change machine part of cycle. Person-machine system
Basic Concepts • • • • • •
Machine maximum time—time machine available during given period Machine idle time—machine theoretically available, but other factors preclude use Machine ancillary time—machine out for cleaning, setting, programming Machine down time—machine out for breakdown, maintenance, repair Machine running time—time during which it is actually running/producing (Trt = 1 – (2,3,4) Machine running time standard—(2,3,4) minimized
3 Indices Machine Utilization Index: Running Time Avail Time 2. Machine Efficiency Index: RT Standard RT 3. Machine Effective Utilization Index: RT Standard Avail Time •
Approach to Improving Restricted Work Maximize 3 indices (approx. 1) Do methods study with machine time in cycle fixed Worker does part of work in cycle while machine stopped (“outside” work) Worker does part of work while machine does its thing (“inside” work) What outside work could be shifted to inside work, thus shortening cycle time?
Allowances in Restricted Work
Personal needs PN – based on total cycle time, worker is at workplace even if he/she not actually doing anything Fatigue FA—based on time (inside/outside) worker actually is doing something productive Try to allocate during machine part of cycle BUT consider whether machine can be left unattended, if not “floaters” may be necessary Most often, personal need times must be outside cycle (cycle time too short to run to restroom!)
Fatigue Allowance (FA) If 90 sec or more unoccupied time, charge it to FA If 30-90 sec unoccupied time, deduct 30 sec and multiple by 1.5
4 Basic Situations All PN and FA taken outside cycle 2. PN outside, FA inside 3. PN and some FA outside, some FA inside 4. PN and FA inside Try to work DOWN this list to achieve # 4 by method study 1.
Unoccupied Time “Allowance”
On some person-machine systems worker is unoccupied during the machine portion of a cycle much more than on other systems Should worker be paid the same when he/she is working vs not working during cycle? Paying same leads to perceived inequities in pay between high-manual vs high-machine jobs Pay rate may be different for “work” vs. “unoccupied” time to compensate
Multiple Machines
1 worker may tend more than 1 machine 2 or more workers may tend 1 machine (e.g., power plant) To study cycle, IE must do timelines for each worker and each machine in cycle “Load factor” = proportion of cycle time required by worker to carry out necessary work during machine process cycle Reciprocal of Load Factor ~ Number of machines worker “could” tend during cycle
Resources Beside the WEB, consider these publications: 1. Groover, M.P. Work Systems and the Methods, Measurement, and Management of Work Pearson Prentice Hall, 2007 2. Kanawaty, G. (Ed) Introduction to Work Study (4th Ed) International Labour Office, Geneva, Switzerland, 1992 3. Mundell, M.E. and Danner, D.L. Motion and Time Study: Improving Productivity (7th Ed) Prentice Hall 1994
