Artificial intelligence and on-chip monitoring in the automotive industry are fueling the drive towards a proactive maintenance revolution and safer roads for all.
The advantage of artificial intelligence to countless industries is in its ability to help humans compile and analyze huge quantities of data to draw actionable insights. When it comes to vehicles, people can generally spot the big issues — such as a smoking engine or a flat tire — but subtle changes in vehicle operations often elude human detection until it’s too late.
What’s Behind The Push Toward Predictive Maintenance?
Identifying early warning signs, such as slight fluctuations in vibrations or temperature over time, could help identify anomalies that lead to failures before they happen. The most ideal situation is not just to manufacture vehicles that can capture these data points, but to create self-diagnosing vehicles that proactively alert drivers before any critical components reach their breaking point. These self-diagnosing cars would extend vehicle lifecycles and even save lives, but the shift towards predictive maintenance in the automotive sector is largely motivated by a shared goal to make reliable autonomous vehicles a reality.
Ensuring Safety Through The Design Of Fail-Safe and Fail-Operational Systems
Any system can fail — automakers know that, which is why designers put a contingency system in place to safely manage failures. For example, if a system failure occurs in a human-driven vehicle, the human will (hopefully) notice the symptoms in time to bring the vehicle to a stop safely. Similarly, in a fail-safe system for an autonomous vehicle, the human driver would be entrusted to take over for the vehicle during a critical system failure.
To that end, Israeli-based software company proteanTecs has designed an on-chip monitoring solution that collects performance data to train machine learning algorithms embedded in their deep data analytics solutions. The idea is not only to provide the data to support a fail-safe system, but to continuously collect and analyze data to catch indications of failures before they happen.
But true autonomy means a fail-operational system is at the helm — which relies on other automated systems to handle failures instead of a human backup. That can be a tough-to-swallow concept for most (particularly regulating bodies), but automotive fail-operational systems are on the horizon. Chassis Autonomy, a Sweden-based startup, has already designed the world’s first fail-operational steering system, and the team is working on a fail-operational braking system too. Here’s how their co-founder and CTO, Thomas Li, describes the firm’s work:
“In autonomous vehicles where a human driver is no longer available as the final line of redundancy; our fault-tolerant, fail-operational steering system ensures steering functionality and availability even in the event of a vehicle or system fault. It will set the new state-of-the-art for critical actuation systems and enable the safe unrestricted operation of autonomous vehicles on the world’s roads.”
Collaborative Efforts Propel Automotive Sector Forward
In order to keep self-diagnosing vehicles safe and financially viable for a large consumer base, collaborations between semiconductor manufacturers, auto companies, and regulatory bodies must standardize a framework for automotive predictive maintenance. The International Organization for Standardization (ISO) recently published the TR 9839 technical report to set the stage for the release of the third edition of ISO 26262, which will underpin these standardizations. Part of this edition includes a functional safety standard through the Automotive Safety Integrity Level B (ASIL-B) certification. Paving the way for future developments in the field, proteanTecs was awarded ASIL-B certification in mid-2023.
Safety is the central component in the widespread adoption of autonomous vehicles. While we’re still years away from self-driving cars usurping human-controlled vehicles, the safety and self-diagnosing technology built on AI and on-chip monitoring continues to advance each year. The days of reenacting strange sounds to your mechanic may be over sooner than you think.
Read more:
● Legacy vs Advanced Chips: The Dance Between Chipmakers And Auto Manufacturers
● What’s Driving the Continued Automotive Chip Shortage