Because the conventional process, due to its high manual effort, focuses on obvious or expected errors, anomalies are often initially overlooked. This carries the risk of subsequent costs and reputational damage – for example, in the case of recalls.
“Our system ‘Sherlock Trace’ takes a fundamentally different approach compared to the conventional process: It proactively leverages the potential of the large volumes of data generated during vehicle development,” says Julian Reindl, Head of Data Science and AI at the ASAP Group. Reindl adds: “With this approach, even sporadically occurring errors are detected. This enables manufacturers to take corrective action before such ‘once mistakes’ make it into production vehicles.”
The ASAP analysis tool can quickly and efficiently identify and evaluate relevant signal sequences. To do this, ASAP experts first model the contextual information for the requirements, before intelligent pattern recognition analyzes these signals and visualizes for users whether they were conspicuous or not. This gives test engineers a clear picture quickly and allows them to derive targeted actions – without having to manually analyze large amounts of data in a time- and resource-intensive manner.
“Sherlock Trace” is characterized by its simple operation via web browser and a flexible system architecture, where the implementation of specific requirements takes place in the backend. This means every use case can be mapped – both project- and manufacturer-independently. Analyses can be performed either on ASAP’s own computing clusters or on customer systems.
“Our tool enables structured, straightforward evaluation of trace data with high process reliability. In addition, together with our parent company HCLTech, we are continuously investing in the further development of ‘Sherlock Trace’ – thus offering forward-looking solutions in view of the constantly changing requirements in vehicle development,” says Maik Ketels, Director Division Electrics/Electronics at ASAP.