How focusing on ‘leaks’ in an OEM’s prototype data flow saved two weeks on every programme globally.
“No-one is going to OWN a car any more”
“We won’t need to do that in the future”
“AI will fix that”.
Our industry is rapidly morphing from building vehicles to providing mobility, as a result, a number of new work-faces have opened up demanding attention. Working out where to place our best efforts is vital.
With the sheer level of disruption from new technology and ideas, it’s easy to assume that everything will change and that improving today’s products and processes is wasted effort. This is not the case. There are many key processes and functions which will be required both now and in the future that we’re moving toward, as well as those which will evolve rather than take one giant, disruptive leap forward.
Though unglamorous, developing these processes should be high on everyone’s priority list, addressing real challenges now and firming the foundations for the future.
CAD to Kit: A Rich Opportunity
Of these challenges, we should be focused on those with the highest value and which are most often used. One such challenge — which has been underestimated and misunderstood for the last 30 years — is the flow of accurate and timely information from design to the production line; or from CAD to kit.
This complex set of transactions involves hundreds of fields of data for thousands of parts, passing through many hands with dozens of hand-offs & interfaces, some automatic, some manual, some paper and many entirely human — and that’s just the initial release, never mind the variants, options and ongoing changes. It should be no surprise that this creates some big challenges or, with the right focus, some big opportunities.
Buckets and Leaks
In our business, we tend to think of this as a complex system of pipework emerging from the drawing board and finding its convoluted way to the production line. It has many valves, restrictors, diverters, filters, holding tanks and a surprising number of people moving buckets to and fro as well as some quite significant leaks. Each individual issue rarely generates sufficient focus to fix it, but the desire to ‘fix PDM (product data management) and ‘implement PLM (Product Lifecycle Management) has been on the agenda for some time.
Why have we failed?
Continuing (and possibly over-extending!) the analogy, a lot of effort and expense has been spent on fixing and improving the valves, filters and restrictors, but the buckets and leaks have continued relatively unchecked. Coupled with the significant increase in complexity — through increased customer choice, more global products and greater use of platform approaches — coping with which has absorbed much of the improvement effort — there is limited evidence of things improving in the last 2–3 decades.
Until we suspend our instinct to reach for the universal big solution (new system, re-org…) and begin focusing on the individual, detailed problems — and the real life human elements of each one — it will be difficult to make progress and easy to rack up big bills.
A recent case study is a good ‘taster’ of what can be achieved inexpensively and relatively quickly when it’s accepted that system isn’t the (whole) answer.
Case Study: Global Prototype Builds
The prototype process is a smaller scale version of the whole programme and a good example of a typical ‘buckets and leaks’ problem. A design is taken from CAD drawing at a point in time, parts are checked, sourced and ordered and they arrive for a vast number of vehicles to then be built.
One major global OEM, suffering from delays to their first production vehicle in successive programmes, found a direct correlation between lateness of prototype builds and lateness of first production vehicle. Further, the biggest cause of prototype lateness was product data issues, specifically BoM issues. As such they decided to focus on BoM quality and the ‘CAD to Kit’ journey for their prototypes and invited Quick Release to join the team.
The approach was two-fold: Tactical support to drive increased performance (plugging the leaks and increasing the bailing rate to strain the analogy!) and then systematic improvement workstreams to underpin this increased performance, such that the tactical support can be reduced and removed.
No new systems were implemented and process changes were limited to those which addressed the specific causes of errors, lateness and omissions. The real effort went into connecting & visualising the information flows and, in particular, the specific human interactions within the flow.
A year in and the results speak for themselves. On this Commercial Vehicle program, we have achieved a higher quality prototype BoM than ever before and parts are being ordered earlier than ever. This has increased the average percentage of parts available at the Material Required Date (MRD) from ~75% in the previous years to over 95% in 2018.
Improving the quality of the client’s product data (in this case, BoM quality), we have seen an increase in the percentage of parts available for every program’s prototype build (P1 to P12).
BoM issues have dropped from the list of problems almost entirely, and all programmes have kept to timing because of it. The latest programme had every single part available on time to its prototype MRD. This can only be seen as a quantum shift in performance, at a crucial time for new product lead-times. And achieved with no AI in sight.