SYSTEMATIC SLOTTING TECHNIQUE
Copyright Edward C Aylward 1986
I should be a multi-millionaire! Back in 1986 I had the idea to develop a complicated statistical maintenance control theory into simple terms. I "pinched" a few ideas and set about de-skilling and simplifying the system.
What came out was SySloTec. I have never used it in practise but have no doubts of it's potential.
I still have the programs and manuals for the implementation.
Anyone out there interested?
The Job Card System
Appendix. I Reference period Data
Append ix. II Run of 'SLOPE' a' 'HI.DATA'
Appendix. III Plot of Ref. Period Data
Appendix. V Run of 'SST'
BACK TO MY BUSINESSES
In maintenance and jobbing work, many work measurement techniques & systems are used. The most common ones are analytical estimating and comparative estimating. In analytical estimating, standard data is used to build up the time for the job. In comparative estimating, a job is compared with a series of other similar jobs and an appropriate time assigned to it. The main problem with both these techniques is that the ratio of estimators to workers is high.
In order to find a solution to the problem of economically estimating the time to do jobbing work, many statistical analyses were made to see if there was a useable pattern. The results were very interesting; there was a discernible statistical pattern to the work. In simple terms, it showed that if the percentage of jobs done was plotted against the log of the times to do the work, then a straight line emerged. This was true in all cases where a sufficient number of jobs were included over a representative period and there was no applicator bias.
SySloTec, (Systematic Slotting Technique), uses statistical methods to set-up an estimating system, but eliminates the need for any statistical knowledge on the part of the user. All that is needed is the input, to a computer, of data for a representative period and the pressing of a button. The monitoring of the application is just as simple.
SySloTec effectively captures the reduced job times brought about by improved methods and supervision. SySloTec, properly applied, controls the drift normally experienced with traditional estimating systems.
With the modern trend towards specialised production, mechanisation and automation, the cost and quality of maintenance is becoming an ever-increasing concern to management. This is of particular importance to the process and mass production types of industry where the cost, and quality of maintenance are major factors contributing to lower costs and competitive production. There are four basic reasons for this: -
1. Mechanisation, automation and computerisation have, directed most companies towards reducing direct labour costs. Indirect costs, including maintenance, have however, increased in both size and importance.
2. With increased automation, breakdowns in production equipment more seriously affect operating efficiency as entire lines or processes may stop.
3. Most companies have long recognised the importance of method study and work measurement as tools of controlling and obtaining the most economical use of direct labour. However, until recent years, techniques have not been available by which a large percentage of non-repetitive indirect work, such as maintenance, could be controlled and/or measured economically.
4. Unions have been reluctant to accept techniques based on methods used in volume production being applied to their work. Management has also been reluctant to face the problems of persuading maintenance and jobbing workers to record their activities for control purposes.
With the use of work measurement and incentives on direct workers the situation arises where a production operative can earn more than the skilled worker.
This is usually dealt with by management paying a 'bonus' linked to production or paying a 'lieu bonus', which has no direct relationship to the productivity of the skilled worker.
Management realises that the higher earnings paid to direct workers are the result of measured higher productivity, and are reluctant to give lieu payments to indirect workers. Nevertheless, they do recognise that genuine pay differentials must be maintained, and that this can only be achieved by extending the methods and measurement techniques used in improving productivity in direct areas.
It has been found many times, that when studied, the uncontrolled, unmeasured; maintenance worker is less than 50% effective. This is compounded by the fact that without proper recording of work done and the lack of good preventative maintenance, downtime is excessively high. When the indirect worker is measured ,and control systems installed, improvements in overall productivity of plant can exceed 150%. Place this improved productivity in the light of the marginal cost of each extra unit produced, (and thus higher profit on each marginal unit produced), and spectacular overall profitability and/or competitiveness can be envisaged.
The two main reasons for not applying work standards and methods to maintenance and jobbing work has been the high cost of measurement and the 'political' problems involved.
1. Job standards set by individual time studies are out of the question. Such a system would require such a high ratio of study-men to workers as to be prohibitively costly.
2. Analytical estimating and comparative estimating can be alternatives but the ratio of applicators to workers is still high. Earnings drift can also be a problem.
3. 'Historical' estimating and 'cost plus' have been used but the lack of accurate recording and the effect on competitiveness has lead to the discarding of such primitive systems.
4. Maintenance management is wary of having systems installed that measure the effectiveness of their departments and give a true analysis of maintenance productivity.
5. Unions have zealously guarded the apparent mystical skills required to maintain complex equipment and claimed that they are not subject to method analysis or measurement.
Most plant maintenance work is characterised by a large number of small and medium sized non-repetitive jobs. The work is normally performed under a wide range of conditions that give a number of variables. The problem is to have a standard that is accurate enough to include the variables but flexible enough to be economically applied.
Accurate measurement of each man or each job is impracticable. It is unrealistic to expect every man to do a set job using exactly the same methods every time he does that job. The conditions in which the job is carried out and indeed the very work content in the job can be variable.
The answer to setting standards for maintenance workers is comparatively simple. The obvious must first be accepted: -
IT IS NOT POSSIBLE TO SET ACCURATE STANDARDS
FOR EACH MAN FOR EACH DAY THAT HE WORKS
IT IS POSSIBLE TO SET STATISTICALLY ACCEPTABLE
STANDARDS FOR A GROUP OF WORKERS BASED ON A
SySloTec, (Systematic Slotting Technique), is based on a simple concept that gives a basis for a practical way to measure maintenance and jobbing work economically. It is based upon the range of time in which a job will probably be done by a qualified person. In simple terms, it can be said that a person given a job to do in say, from two hours to four hours, will be within that time slot 95 times in 100.
SySloTec has been developed, using standard statistical methods, to set standards for maintenance and jobbing work that will give an accuracy of +/- 5% at 95% confidence levels when applied to a group over a representative period. The statistics necessary are contained within the computer program. All that is required is a rated activity sample of jobs over a period to generate the slots needed to apply the technique. Checks made each six months ensure that the system is capturing method and organisational improvements and that no 'drift' has occurred.
An analysis of the work done by maintenance workers will most likely show that some 80% of jobs require less than one shift to complete. Also, longer jobs can be split logically to bring them within that time range. Long jobs not split can be dealt with separately.
Highly repetitive jobs are given a discrete time and included in the system.
From the initial activity sampling study data, SySloTec develops a tailored system using five to nine slots, (usually between five and seven are sufficient), to cover job times of between fifteen minutes and twelve hours.
All that the applicator has to do, is to say, "Will he do this job in less than 1/2 hour? Of course not! Will he do it in less than 1 hour - doubtful. Will he do it in less than 2 hours? Yes, he will." The applicator would then use the slot with the mid-time of 1.5 hours.
The types of work done will vary in different firms. But, the basic principle is true, that the slot should be wide enough to reduce the doubts of slotting, yet narrow enough to give a statistically correct measurement of the work in the period. SySloTec achieves these objectives with the minimum of calculation to the user.
The allocation of men into self-identifying groups will be a matter of custom, tradition and practice. Rarely will it be beneficial to split naturally occurring groupings in order to satisfy the requirements of the system. Where very small groups do occur the measurement period is extended, and if payment by results is to be made, a two to four weekly running average may be used.
The slotting of the jobs is done by people having a good knowledge of the actual work content of the job and the conditions under which is to be done. They must be experienced artisans.
It may seem that the use of the range concept would make the standard for each job so inaccurate that the overall accuracy would be completely unacceptable. In practice, this is not true. The SySloTec modules, though simple in presentation, are based on well-tried statistical methods and are therefore acceptable in the 'real world'.
There are six requirements in using SySloTec: -
1. Introduction and installation of the system must be pre-planned in close co-operation with management, supervision, and unions.
2. The introduction and/or the verification of the Job Card system.
3. Formal organisation of the work load. (inspections, servicing, planned maintenance, breakdowns)
4. The supply of adequate tools & materials and good supervision.
5. Rated activity sampling.
6. Well qualified and experienced applicators.
NB. At all times complete honesty and openness must be maintained. The motto should be :-
GIVE AS MUCH INFORMATION AS POSSIBLE AS EARLY AS POSSIBLE*
There must be mutual trust between management and workers.
Great benefits can be derived by the installation of SySloTec and proper control procedures with or without a linked pay structure. The arguments for and against payment by results are a matter for management and workers to discuss.
As the system settles in and the control procedures become effective, the time to carry out jobs WILL decrease, and the slots WILL be adjusted. This should be reflected in the reduction in man-hours to maintain the plant satisfactorily. If it does not then the likelihood of bias and drift is investigated.
THE JOB CARD SYSTEM
The biggest hurdle to overcome in introducing a systematic method of maintenance is the use (and abuse!) of the Job Card System.
If a Job Card is in use and has not been subject to strict control then one or more of the following will be evident:
1. Job times are in multiples of shifts, half shifts or at best hours.
2. No unoccupied time or waiting time is recorded for control purposes.
3. No detailed job instructions.
4. No reference to tools or materials required.
5. Job Card information is not analysed after completion to update plant records.
6. Authorisation for work is not centralised.
7. Work done is not checked and signed 'OK' by supervision or other authorised person.
Before introducing SySloTec, it is imperative that the principle of a controlled Job Card system is accepted.
The specific design of the Job Card should be done in collaboration with all who are likely to use it or the information it contains.
It should contain at least the following: -
1. The authority for the job to be done, (works order, cost code, cost centre etc.). The authorising person must sign the card.
2. Plant reference number.
3. A brief but clear description of the work to be done. The description should be specific and not general.
4. A list of any special tools required.
5. A list of parts required.
6. A reference to any safety precautions to be taken.
7. The ACTUAL time taken to do the job. All ineffective time must be logged. This is done by the worker, but it is the supervisions' job to ensure that correct times are recorded.
8. An authorised signature to confirm that the work has been carried out satisfactorily.
The issue of the Job Cards should conform to the following outline procedure Again the actual detail must be decided by the people who will use the system.
1. Job Cards must be issued by the level of supervision directly controlling the person who is to do the work. This ensures that the best man for the job does the job.
2. If a job is not completed then the card must be handed back to supervision and the reason recorded.
3. Several Job Cards can be issued to one person at the same time if the total time covered is, say, less than half a shift. If a man is working at a distant location, he does not have to go back to the office several times per hour'.
4. Several jobs can be entered on one Job Card. e.g. clean, check and replace 24 light fittings in cost centres IA, 2D and 4C.
5. When the job is complete and signed off by the supervisor the Job Card must be forwarded to a central point so that information can be transferred to the Plant History Record and frequencies revised. This is a key step in the system. It is at this stage that the information is analysed and decisions made regarding the future. Planned Maintenance inspection, repair and replacement schedules, thus minimising breakdown and emergency maintenance
6. The Job Cards should be passed to the relevant sections for cost allocation and any other analysis.
The first step, after the initial induction talks, is the implementation or checking of the Job Card system.
It is possible, though not desirable, to start the implementation without the Job Card system, but the total system cannot function without good Job Card control.
If a good booking system is not in practice but job times are recorded then an activity sample could be taken just to measure the effectiveness (rating) of the workers while working. This rating could be applied to the booked times in order to give the 'standard time' for the jobs.
Ideally, a complete rated activity sample should be taken to compile true standard times for the jobs during the study period.
The representative period is decided - preferably a pay period of say one-week and the number of jobs in the period is ascertained. The number of jobs should be between 150 and 400. In order to get this number of jobs it may be necessary to cover two, three or even four pay periods if the group is very small. If pay is to be linked to performance then a moving average should be used to smooth out any anomalies in the work pattern.
The rated activity sample is carried out over the representative period. It should be remembered that a true picture would not emerge until the sample period is well under way. It is, therefore, recommended that the sample period cover two three or more pay periods in order that the beginning is not unduly affected by the over-awareness of the workers being studied.
The job tine data is collated (APP.I) and entered into the SySloTec module SLOPE~. This gives an output of factors and percentages (APP.II). These factors and percentages are plotted on a graph (APP.III). The line of best fit should approximate to a straight line.
If not it is most likely that not enough readings have been taken and further readings should be entered. When assessing the line of best fit do not be too concerned about the top and bottom tails. Providing the line is OK between 15% and 851 then SySloTec can cope.
It is likely that the line will be straight between 15% and 85% but a 'peak' will occur at a particular point. This will be due to the booking, or setting, of job times to coincide with a clock work period; e.g. shift, half shift or to the nearest break. If this happens the 'SySloTec' module 'HI.DATA3 is used (APP.II). This will ask for the factor under query and give the time period to investigate.
If after the readings for a representative period covering 400 jobs does not give ~n acceptable straight line; then either the whole job time booking system is at fault or the job times do not conform to the usual distribution. The latter is most unlikely.
When sufficient data has been collected to give a straight line the data is used by the SySloTec module 'SST' (APP.V). This module gives the slot reference number, slot time upper boundary and mid-slot time.
The times are in hours. Times are not entered on the Job Card, just the slot reference number. The slot reference number must not pre-printed on the Job Card as this can change as the system captures the effects of method and organisational improvements.
In slotting a time, the applicator decides the lowest slot time boundary below which the worker will be able to complete the job. Job times in excess of the highest slot must be split or excluded from the system. The mid-slot time is used to calculate performance indices and if required, used as the basis for a payment by results scheme.
In the early stages sets of data collected during the set periods should be entered into the SySloTec modules 'SLOPE' and 'SST' to check if the times are conforming to the reference period data. If not, then further analysis must be made. The output of the system should not be used for payment purposes until a 'steady state' has been reached.
When it is evident that the system is operating consistently then the slots should be updated at about six monthly intervals. The effects of the Job Card system and SySloTec should show in the gradual reduction in the number of standard hours required for a set number of jobs and a reduction in breakdown and emergency maintenance. The applicators must be made aware that as supervision and job control improve that the job times given for set jobs will reduce.
SySloTec effectively captures the reduced job times brought about by using improved methods. SySloTec, properly applied, controls the drift normally experienced in traditional job estimating systems.
0.4 0.4 0.9
0.9 0.8 1.1
1.1 1.2 1.2
1.3 1.5 1.4
1.2 1.4 1.5
1.6 1.6 1.6
1.5 1.8 1.8
1.8 1.7 1.7
1.9 1.9 1.9
1.8 1.7 2.0
1.9 1.8 2.0
2.0 2.1 2.1
2.0 2.0 2.1
2.4 2.1 2.3
2.2 2.5 2.4
2.6 2.4 2.7
2.5 2.7 2.6
2.6 2.9 2.8
2.8 3.0 2.8
3.0 2.9 2.9
3.0 3.0 3.0
3.3 3.4 3.4
3.4 3.3 3.3
3.5 3.4 3.6
3.7 3.6 3.6
3.6 3.9 3.6
3.9 3.7 3.8
3.8 3.7 3.6
3.8 4.1 4.7
4.7 4.1 4.8
4.3 4.1 4.9
4.5 4.5 5.2
5.0 4.6 5.3
5.6 5.1 5.3
5.1 6.1 5.7
6.0 6.0 6.3
6.4 6.1 6.0
6.9 7.0 6.7
7.3 8.1 8.0
8.0 8.1 9.2
8.8 9.4 8.9
9.8 10.2 10.1
10.9 11.9 11.7
12.5 12.5 12.7
14.5 14.4 16.7
19.6 24.6 30.3
40.1 41.7 2.3
1.7 3.1 4.6
4.6 5.2 6.7
8.5 7.9 9.2
OUTPUT TO SCREEN CS) OR PRINTER (P)? S
FILENAME FOR INPUT? SI
RUN "HI. DATA"
FACTOR OUT OF RANGE? 26.2
CHECK TINES AROUND 13.7 HRS
OUTPUT TO SCREEN (S) OR PRINTER (P)? S
FILENAME FOR INPUT? S1
NO. OF JOBS IN REPRESENTATIVE PERIOD? 400
SLOT UPPER MID
NO. TIME TIME
1 3.29 1 64
2 5.96 4.63
3 6.74 7.35
4 12.51 10.62
5 17.99 15.25