The modern automobile has several ECUs each of which receives inputs from various sensors and sends commands to multiple actuators. Further, it communicates with other ECUs of related subsystems. Some implement performance critical functions such as fuel-injection and ignition-timing, whereas others control safety critical functions such as Anti-skid Braking (ABS) and Electronic Stability Control (ESC). An automated manufacturing test station for the ECU is therefore significantly complex in design, involving several pieces of instrumentation, sensor simulation and multiple communication protocols.
Take a mid-size OEM, for instance, that sells around 100,000 vehicles annually. Considering engine control function alone, it means that their tier-1 vendor needs to supply (at least) an equal number of ECUs. The manufacturing rate would be around 8 ECUs per hour, assuming 24x7 operations in 3 shifts, for which four parallel assembly lines would imply less than 30 minutes to manufacture 1 ECU!
The time budget for testing at the End-of-Line (EoL) is even shorter. The operator would get less than 1 minute typically, including loading, executing automated tests, knowing pass or fail, printing and affixing bar code for passed unit (or placing failed unit into reject/rework bin), unloading and ready to load the next unit! Different versions of the ECU that are simultaneously in production bring on additional complexity. The operator would have to reconfigure tests each time for a different version, typically within 4 minutes.
Let us now examine where this challenge could be acute in the industry versus where it might be non-existent. Any tier-1 manufacturer (or OEM) already in the supply-chain would already have solved this challenge in their factory floor and if not, then they would hardly be selling! However, a new ECU design just getting into production may not enjoy a similar ‘steady state’, be it part of a new vehicle brand that an OEM plans to introduce or be it related to an additional feature that is being introduced for the first time for a new variant of a model, like, adaptive cruise control.
In these scenarios, does the tier-1 manufacturer, while in the process of taking the new design through the NPI process, have the required bandwidth to design the test station as well? In the case of technology transfer of ECU design from a global principal, does the manufacturer have in-house expertise in the early stages to develop a test station on time before pilot production starts? In the alternative scenario of in-house development, does the manufacturer really have the resources, bandwidth and the timeframe to design and manufacture the test station before the design passes all the type tests and hits production? Alternatively, do their existing test station vendors for other components, like starter motors, tilting mirror assemblies or instrument clusters have the necessary expertise to design such a complex test station?
What about ECUs for the upcoming wave of electric vehicles (and hybrids) that are predicted to transform the entire motoring landscape forever? Not to forget the twowheeler (and three-wheeler) segments, which under the rapidly closing time window of emission control regulations (Bharat Stage-VI in India has a 2020 deadline currently!) will be forced to switch to ECU-based engine control in a few years’ time in order to continue selling legally.
Design reuse for ECU test stations
Here’s where some visibility into the design of ECU test station provided insights into accelerating its development. Certain parts of the test station always remain very specific to the particular ECU design, like the ECU’s connector for which the test station provides a mating connector. However, considerable amount of functionality remains common and is very generic, for instance, the Human Machine Interface (HMI), which is the main operator visible part that he sees and operates continuously. Another example is inter-ECU communication capability over multiple automotive bus standards and messaging protocols.
Expertise in vehicle diagnostics, testing and simulation techniques enables modular development of common functionality (hardware and software) as a generic test platform. The design of a test station for a new ECU is, therefore, reduced to customising the platform for the design specifics of that ECU. Mapping the test platform to the custom requirements of the specific ECU along with competence in digital/mixed signal hardware design enables quick customisation of specialised parts of the test station, thereby, accelerating cost-effective development.
Panel push-buttons to start automated test sequence, lamps to display test-in-progress status and hooters/lamps to alert failed tests make the operations intuitive to operators. The mounting, orientation, peripherals for viewing/printing and display properties are all ergonomically designed, optimal for continuous usage by operators. Routine and rough/careless usage by operators constitutes a really harsh environment especially for parts such as the ECU’s mating connector, so the design ruggedness has to stand the test of months of continuous loading and unloading operations.
Flash it first, then test it
Another closely related area for manufacturing test efficiency is ECU flashing. Once an ECU completes electronic assembly, operators first use a flashing tool to download firmware into it without which it cannot be tested. The design of ECU flashing unit is similarly accelerated by a generic flashing (hardware and software) framework. The only design input required from the manufacturer is the algorithm for unlocking the ECU for flashing. The customer’s (or principal’s) confidentiality of this critical algorithm could be protected by including it as a library (in binary form) so that source code containing proprietary logic does not get revealed.
All of it comes together finally in the hands of the operator, who after loading an ECU has less than 1 minute to run the tests to know if it is a pass or a fail. Pass is good news always as the ECU gets a bar-coded label stuck on it and moves forward to the next stage. However, fail is hardly the end of the road. In order to keep the rejection costs low, failed units need to be repaired without delay for which the test station provides precise troubleshooting assistance.
As has been proven at multiple manufacturers’ already, differentiation between generic/re-usable versus specialised parts of the test station design avoids the need for re-inventing manufacturing test station for every ECU design. However, complete assurance of the customer’s IP is an implied and absolute requirement for the test station designer to honour. Finally, experience and proven track record of installing ECU test stations on factory floors and supporting production personnel in the usage and fine-tuning of these systems are prerequisites to ensuring efficient and trouble-free operation for the entire production lifecycle of the ECU.
The article is authored by Ram Mohan R K, Vice President, Deep Thought Systems