Well, no and no.
A first tier auto supplier was struggling to meet demand from a particular line with OEE's running at around 50%. It had decided to invest about £40,000 on a new automated end of line inspection device since its management information was telling it that was where the majority of downtime was originating (which indeed was correct at face value).
However, process mapping and going to the Gemba (workplace) revealed that a much bigger problem was being caused by the work content at the three semi-manual assembly stages. In this case, it was not 'Rocket Science' to understand the reduction in average line speed was being caused by uneven work content and lumpy flow. As is so often the case, the lowest skilled (agency) workers were at this stage despite it being the bottleneck.
The work content was assessed using video and content analysis and the tasks at the three stations were radically changed. A trial was done showing incredible increases in OEE (towards world class 85%!) and the team leaders at the company quickly installed mistake proofing for the new operation (without prompting, we hasten to add).
The value to the company in increased output was over £700,000 per annum.
OK, it may be rare to find such amazing increases in throughput for a relatively small resource input, but ask yourself;
- Are there any semi-automated lines in your workplace where the data you are collecting is actually leading you to the wrong conclusions?
- Is the bottleneck in a different place to where you suspect it to be?
- Have you been to the Gemba and understood the reality?
