Coursework 1

Coursework 1

Background

The company manufactures a range of specialist radiator and heat exchangers. A Key component of all products is the metallic core tubing. Every product range contains these items. The company produces these tubes on a single machine. Due to the criticality of this equipment it is important to increase its effectiveness.

The process itself entails two raw material types, depending on product (brass or tin). Reels of raw material are loaded to the rear, and fed through the machine. The raw material is then folded, welded and cut to the required length, so producing the finished component.

Current situation

The company has designed a data collection chart and instructed operators to collect data. Currently there is one week’s worth of data collected. No corrective action has been taken.

All breaks are covered (the machine is planned to run continuously)
The shift starts at 8am and finishes at 4pm
A set-up is defined as a change of material or cutters (or both)
Speed is recorded digitally and reported as an average over the shift.

1 Calculate the daily availability, performance, quality and OEE figures (state any assumptions made).

2 Pareto analyse the losses.

(Bonus marks may be awarded for working out what the “Shift M/C speed” column really indicates, if anything).

Remember – this is an individual piece of coursework. If you collaborate, you risk losing all your marks!

Note – this is A model answer, not THE model answer, as so much depends on the assumptions made, especially the rated speed (for the performance calculations)

A series of assumptions need to be made before the problem can be tackled. These will include (but not be limited to)

The units of measure:
Product – two product “families”, brass and tin.
Length good – number outside brackets = length of each good product in mm – Number inside brackets = quantity made.
Therefore, length good = both numbers multiplied together Length bad – length of bad material cut out? measured in mm? inches?
START and FINISH – time that batch was started and finished on the machine, clock time.
Minor stops, Start up Losses, Reel Changes, Material, Breaks, Breakdowns and Setups – all indication of lost time due to each cause. Measured in minutes.
Reel changes also has a number in brackets – number of reel changes completed.
No of Operators – appears to be unused
Comments and Maintenance – a record of exceptional events? Shift M/C speed – unknown units.

Other assumptions – Breaks are supposed to be covered, therefore no planned downtime? But what if policy is being ignored and breaks result in the machine stopping?
Operators are recording data accurately and consistently
Speed for batch with unrecorded time on Friday can be inferred from speed(s) on similar products earlier in the week. (of course, for OEE calculations this is not necessary, all we need is the 8 hours worked and the downtime records)
If feeding the tube through the machine is the critical, performance limiting process:

A rated speed will need to be assumed to allow Performance Calculations to be made. The student should explain the reasoning behind their assumed rated speed.
(Example calculation:
We know the claimed start and finish times for each batch. Let us assume that the machine works continuously (i.e. no stops) on each batch, and that the batch contains no bad lengths. We can then work out the speed of the machine when producing each batch.
Speed of machine = total length of batch (length of each piece x quantity in batch) elapsed time
Example – first batch on Thursday, length 1347mm, quantity 341, Elapsed time 30 minutes.
Speed of machine = 1347mm x 341
30 minutes
= 459327mm
30 minutes
= 15310.9mm/min or 15.3109m/min)
The “fastest batch” was the third batch on Monday, which was produced at a speed of 24.7728m/min. This is unlikely to be the actual rated speed of the machine, and knowing that machines tend to be specified in round numbers, a rated speed of 30m/min seems a reasonable assumption.
Alternatively, cutting may be assumed to be the critical process, in which case the length of units is not necessary, just the quantity, and “length bad” may be assumed to be products of unacceptable quality isolated after the process.

Calculations (see table below)

TD TUBE MACHINE O.E.E. DATA
Monday Brass 1347 786 1058742 25/24/10
25/5/4 08:40 09:40 60 17645.7 17.6457 1058.742 5 – 40 51 9 25 45 18 26 Bad cutting finish on tube, knives changed 18 30000
1283 1028 1318924 09:40 11:00 80 16486.55 16.48655 989.193
1588 780 1238640 11:00 11:50 50 24772.8 24.7728 1486.368
1588 2350 3731800 11:50 15:30 220 16962.73 16.96273 1017.764
1283 300 384900 15:30 15:55 25 15396 15.396 923.76
Daily total 7733006 2362 435 17777 17.8 1067 56.3 62.1 95.5 100.0 53.7
Tuesday Brass 1283 708 908364 5/15/9
4/6/40
6/8 08:00
08:45 45 20185.87 20.18587 1211.152 5 – 66 11 21 40 20 32 Bad cutting finish
Dross out solder
Changed flux
Set up machine for tin
19.5

1588 3156 5011728 08:45 13:05 260 19275.88 19.27588 1156.553
1347 1935 2606445 13:05 15:20 135 19307 19.307 1158.42
Daily total 8526537 2362 440 19378 19.4 1163 61.7 67.3 96.0 100.0 59.2
Wednesday 1347 1836 2473092 35/20/10/
6/22/10 15/20/6
15/20/10
6/15/25 08:50 15:10 380 6508.137 6.508137 390.4882 – 60 9 29 50 14 19 M/C down
Bad finish on tube Removed knives and cleaned
7.2
Brass
Tin
1588 17 26996 15:10 15:45 35 771.3143 0.771314 46.27886
Daily total 2500088 5969 415 6024 6.02 361 64.2 71.6 27.1 99.8 17.4
Thursday Brass 1347 341 459327 6/10/6
10/6/6
6/8/29 08:00 08:30 30 15310.9 15.3109 918.654 1 – 48 8 16 30 20 29 Marks on tube
Pick up on tinned brass
17 15
1941 1194 2317554 08:30 11:00 150 15450.36 15.45036 927.0216
1283 1082 1388206 11:00 12:20 80 17352.58 17.35258 1041.155
1347 2040 2747880 12:20 13:40
14:40 140 19627.71 19.62771 1177.663
Daily total 6912967 2210 400 17282 17.3 1037 70.0 74.7 68.6 100.0 48.0
Friday Brass 1347 2808 3782376 40/15
6/6 08:00 13:20 320 11819.93 11.81993 709.1955 1 – 70 12 19 75 14 39 Lost time during set up 15 17
Tin 1588 960 1524480 12/6/10/6
6 13:20 15:55 155 9835.355 9.835355 590.1213 Greased rollers
Daily total 5306856 2717.8 475 11172 11.2 670 54.6 64.7 67.6 99.9 36.9

Weekly total 30979454 15621 2165 12 40 295 49 110 240 86 145 61.3 68.1 70.2 99.9 43.0

The Pareto principle is named for the Italian economist Vilfredo Pareto, who observed that in Italy at the start of the 20th Century, 80% of the land was owned by 20% of the population. The business and quality guru, Joseph Duran took this idea and expanded it, suggesting that in many instances found in the business world, approximately 80% of the effect can be explained by approximately 20% of the causes. So, if we can tackle the “vital few” problems we can have more impact than if we try to tackle the “trivial many”.

In our case, the first order problem is NOT quality, and whether it is performance or not depends entirely on our assumed rated speed. We have the “facts” for the causes of poor availability, so it makes sense to analyse these. It makes sense to do this analysis for the whole week rather than daily, because this way we can avoid getting distracted by unnecessary detail of issues that may be “local” to one day, and less relevant over the whole week.

The Pareto Analysis shows us that in this case the “vital few” causes of lost time on out TD tube machine are:
Reel Changes
Breaks
Set ups
Between them, these three causes account for over 70% of the lost time recorded.

Reel changes may be tackled by possible buying longer reels, or perhaps by Quick Changeover techniques, Breaks are a result of the company policy being broken, so this is relatively easy to tackle – ensure there is skilled cover available whenever the operator has a break scheduled.
Set ups, like reel changes may be tackled by Quick Changeover techniques.