IAV Featured Story Tracking down errors

Tracking down errors

Vehicle connectivity is in flux, shifting away from conventional systems and toward complex network architectures. This poses huge challenges for vehicle manufacturers with everything from data collection to validating the full vehicle. IAV offers a comprehensive big data value chain and provides support for testing all electrical and electronic components with innovative methods (data analytics) and proprietary systems.

From real-time traffic jam information to distance warning systems and automated search functions for parking spaces, today’s vehicles have a variety of smart features that offer true value for occupants. At the same time, connectivity is on the rise: Complex
assistance systems, mobile online services, and communication between vehicles and their surroundings creating a constant flow of information. For example, various functions and the communication involved in the Car2X segment or driver assistance systems and AR (augmented reality) applications are growing increasingly complex – especially when it comes to validation and testing. Huge volumes of data on a vehicle’s movements, status, and surroundings arise. For vehicle manufacturers, these unstructured volumes of data – big data – offer tremendous advantages, but also several challenges. One of those challenges is the connectivity architecture used for vehicles. Similar to the network architecture found in the IT world, this architecture has to cope with large amounts of data.

So far, conventional systems such as Controller Area Network (CAN bus) have been used, which are designed for a low data transmission rate and a neat information structure. These systems are now replaced by the Automotive Ethernet communication standard and by network architectures based on high- performance computers. Thanks to their tremendous processing power, information can be requested from certain control units ad hoc, simultaneously and in parallel. All this makes for a complex network architecture waiting to be mastered.

«OEMs are taking new approaches to meet the challenge. We help them to validate the new network structures and make them transparent using both proven and innovative analytical methods.»

Richard Benedix — Team manager for data analysis and reporting

Network structures and data analysis

Automotive manufacturers require deep insight into the world of big data. IAV helps them visualize what computer scientists program and engineers develop. Interactive dashboards and structured reports allow OEMs to trace flows of data and network communications throughout the entire system.

To achieve this, IAV has created a holistic data analytics value chain that encompasses all of the services needed to analyze measurement data, from data recording to preparation and interpretation and to visualization. “Companies can use these data to track down errors and make solid decisions – we give them the tools to do it,” Benedix says.

It starts right away with the documentation of the data in the vehicle. Modern measurement technology is used to determine which information should be recorded, when, and in what formats. Vehicle connectivity has changed a lot in this regard as well: Concepts for logging measurement data have been adjusted with an eye to performance and data recording so that the volumes of data generated – which are unusually large by past standards – can be processed efficiently. Data that used to take up just a few gigabytes (GB) of storage, can in some of todays test drives reach up to 200 GB each.

The data are stored in the measurement data platform, which gives OEMs a comprehensive management tool. The Web-based front end gives them a full overview of all vehicles, measurement devices, and analysis. Because the data are transmitted in real time, they can even perform analysis and troubleshooting while a test drive is in progress.

Fehlersuche und Reportings

The interactive data features allow even users with no programming skills to browse for certain events in their data, using simple search language. Machine learning algorithms support the process of searching for possible errors by identifying anomalies in the data patterns.

An automated measurement data analysis feature (AMeDA) shows customers complex contexts surrounding various errors. To this end, data scientists and data analysts from IAV generate reports in line with customer questions, programming templates that visualize certain matters in diagram form. After that, the customer can pull the report automatically from the measurement data platform. AMeDA and interactive data analysis are highly flexible, as these systems can be adapted individually to any data
format or data logger. Both tools are also distinguished by their scalability and expandability, which makes the entire tool chain future-proof. Even if vehicles generate zettabytes of data, the interactive data analysis and AMeDA process the data quickly,
without any issues. In this way, the IAV value chain harnesses rapid automated analysis and direct visualization on dashboards to unlock direct findings to resolve errors and provides the openness and flexibility needed for future-oriented engineering.

The article appeared in automotion 03/2021, IAV’s automotive engineering trade magazine. Here you can order the automotion free of charge.

What moves us

Brain & heart – portrait

Placing the focus on people

Following her dual degree in computer science, Sara wrote her master's thesis at IAV. She is now driving software development and connected services usability forward as a team leader. She is comm ...

Stories

How to talk to charging stations

It is essential to utilize the power grids more evenly with intelligent and flexible charging control of e-vehicle fleets. This not only avoids bottlenecks in the power supply, but can also lead t ...

Stories

Stop hackers

Cybersecurity: In the future, no car will be on the road that has not been developed to be cyber-secure. IAV helps OEMs and suppliers to cope with the new requirements and processes.