The figures in these studies come from people inside those operations – engineers and managers who dug the savings out of their own data. They are not audited or officially approved; companies do not publish numbers like these. That is why we do not name the customers, and why every figure says where it comes from.
The ventilation plant was sized for a fully occupied hall, yet production fluctuated far more – ventilation ran on a fixed schedule regardless of how many people and machines were actually working. The filters on a production machine, meanwhile, were replaced on a prescribed time schedule even though the machine did not run continuously.
Aisle lighting ran at full output for the whole shift, even in aisles nobody entered. The fittings were the halogen type common at the time, drawing hundreds of watts each.
The electricity saving came from two steps at once: swapping halogen fittings for LED cut the power draw roughly tenfold, and motion-based control added a further few per cent on top – lights run only when a person or equipment is in the aisle.
Consumption readings were scattered across systems that could not see each other, and reports were assembled by hand.
The size of the forklift fleet was based on an estimate rather than on how much the trucks were actually used.
The measured utilisation showed that the planned purchase of two more forklifts was unnecessary – and after adjusting processes the company managed with one fewer forklift than before.
The company had no real-time picture of production. Plan and actual output were compared only after the fact, which also held up dispatch planning.
The performance and quality components come straight from the controller, and the operator-entered downtimes determine availability – together the full OEE, plus order fulfilment.
Purely manual production of unique pieces – no automation to read data from. The only thing measurable was the number of finished pieces per shift; workstation times, inter-operation times and downtimes were invisible.
The times between scans yield the OEE of every workstation, of the logged-in worker and of the whole line: actual times are compared with the standard times for the given product type, planned downtimes (cleaning, testing) with their own standards, and unplanned downtimes determine availability.
A compressed air leak is neither visible nor audible over plant noise. Compressors make up for it with higher output – electricity consumption and compressor wear both rise.
The company located the leaks that way and fixed them.
Internal, warehouse and customer labels did not match. The error surfaced at the customer – a complaint and a contractual penalty.
Faulty deliveries are caught before dispatch, not at the customer.
We haven't published a study in this area yet – we'll show you one at the meeting.
30 + 30 minutes, no commitment: an introduction to the system, then a demo on a problem of your own.
We measure traffic with Google Analytics – only if you allow it. More in the privacy policy.
