This case study is fictional. However, individual events presented are tiny drops of experience.
A manufacturer of industrial machinery with a long record, but without its in-house electronics development, requested our services. The product in question involved:
- Middle power actuators
- Buttons and touch TFT
- Safety features
- Sensors
- Signal processing options
- Communication
- 30k pcs/year for six years
The customer set the budget for the development and testing of the electronics delivery at 900kEUR based on their experience. We considered this target low, but we accepted the challenge. We agreed to a 90kEUR fixed price for our services, the first fourth due before release of the RFQ.
During RFQ preparation, we were part of the RFQ team and performed a final review of RFQ documents.
Some of our suggestions:
- We defined degraded operation parameters for safe states and elevated temperatures. Otherwise, the product would be unsatisfactory even if up to specification. Savings: up to 10eur/piece and 100kEUR and 6 months in change management.
- We suggested doing in-field measurements to specify EMC immunity. General industrial EMC norms reference only set deficient specific immunity levels definition. Savings: redesign based on field failures for up to 300kEUR, up to one year.
- We included SW tests and SW release plan requirements. Savings: up to 200kEUR, up to 6 months.
- We suggested safety analysis requirements.
Even after this tedious work, the customer still needed some time to define the product fully – we are still keeping our hopes to see a project without change management.
The customer piece price target was roughly evaluated as well and deemed reasonable.
One company fitted in the targets. The supplier suggested mixing human-machine interface and signal and actuator tasks on a single MCU. We enquired the supplier about further details of his architecture. The customer did not nominate the supplier because it could not answer questions regarding real-time tasks in its hybrid architecture. Savings: up to 300kEUR, up to 8months
The second company exceeded the target by roughly 200kEUR, including license fees for safety-related libraries and 10eur/pc. This supplier did its homework, and we found no significant issues during the review of its offer; hence it was nominated.
The customer dropped, changed, and introduced some requirements during concept preparation. We actively participated in this process. It also enabled the design to cost opportunities. The control boards could now utilize ready-made MCU boards, and several actuators can be used for shared outputs. The supplier did not propose these savings alone but agreed to them. Savings 12eur/piece, 200k in development – we reached the targeted value.
During reviews, we made enough suggestions to improve the maturity of the electronics by at least one sample loop. Savings: up to 50kEUR
We asked the supplier for proof of allocation of essential components. The development supplier hesitated a lot and then provided allocations short by one month into serial production. The customer accepted this, some lead times extend over one year these days.
The supplier consulted us with test plan deviation suggestions. We approved some and rejected some which did not reflect customer needs. The supplier was also backed by us while proposing sigma-delta ADCs instead of SAR converters, and we performed an independent simulation to prove they were ok to use. Savings: none, only better design and tests
The development supplier was also the manufacturing supplier, so we did only a brief manufacturability review.
Pre-series prototypes failed surprisingly during EMC qualification. Some of our suggestions proved helpful—savings: up to 30kEUR in testing time.
Functional and environmental tests had two findings we, together with the customer, approved as acceptable deviations.
We made suggestions for end tests to reflect customer needs more tightly. Savings: well, our recommendations increased the price of manufacturing by 2eur/pc (increased scrap)
The customer made the final payment for our services two months after the start of production as agreed. Occasional field failures are consulted with us nevertheless. Our main suggestion would be to replace switches with sulfuration-resistant types, even though there should be no need for them in theory.