The need for proactive and predictive maintenance.
In today’s automotive landscape, the integration of advanced software and hardware has transformed vehicles into complex data-driven machines.
Sensors like cameras, radars, and lidars constantly monitor the vehicle’s surroundings, feeding data to electronic control units that enable advanced driver assistance features like adaptive cruise control, lane-keeping assistance, and collision avoidance. Within the heart of the powertrain, electronics manage the driving dynamics of the vehicle while optimizing performance and efficiency. Moreover, these electronics extend to the infotainment system, providing entertainment, navigation, and connectivity services. From touchscreen displays to voice-activated commands, these systems have transformed the driving experience, turning cars into smart, interconnected hubs.
This shift has led to vehicles being likened to “data centers on wheels,” ushering in a new era of challenges and opportunities.
The New York Times reports that modern vehicles can easily have more than 3,000 semiconductor chips inside. [1] Over time, the electronic systems within automobiles have evolved into the most intricate component of vehicle architecture. Advanced system-on-chips (SoCs) are populated with billions of transistors, a figure that has doubled annually over the past three decades. [2] Furthermore, as the industry shifts to software-defined cars, the rapidly growing software complexity dramatically widens the gap between the industry’s needs and capabilities. Software in today’s cars can contain more than 100 million lines of code. In comparison, software for a modern commercial aircraft contains “only” ~10 million lines of code. The integration is particularly complicated for car OEMs because every car model uses many ECUs on which different software runs. For a whole car, a variety of software running on different CPUs must work together. Operating systems must be orchestrated, reusable (across vehicle lines), and embedded.
Addressing this complexity is imperative for automotive OEMs to ensure reliability, safety, and customer satisfaction. Yet the sheer number of electronics brings an increasingly high risk of system failures. Testing today ends at product deployment to the field, and is usually based on pass/fail criteria, often lacking much needed resolution. When these failures inevitably emerge, swift identification of their root causes—whether stemming from faulty hardware, software components, or other issues—becomes paramount.
As the electronic technologies of software-driven vehicles evolve—pushed even faster by the coming of electric vehicles, advanced driver assistance systems, and autonomous driving—monitoring and diagnostic practices need to keep up. This evolution is essential to meet the expectations of drivers, automakers, and fleet managers, all while mitigating the rising costs associated with recall actions, which presently amount to a staggering $15 billion annually. [3]
The key to preventing failures (the inability to perform properly) is to identify and resolve faults (abnormal behavior or defects) before they become failures. It’s important to note that diagnostic practices detect faults only after they’ve become permanent, and by then failure may be just seconds away. Typical descriptive and diagnostic analytic systems can’t always pinpoint the root cause of a problem, which leads to time-consuming and costly trial-and-error processes.
In contrast, advanced predictive and preventive maintenance methods predict what will happen in the future and present the best way to prevent problems from happening. They leverage sophisticated sensors that implement many data points as high-quality performance indicators. Analyzing that data can reveal when a system is in a critical state. Alerts are generated days, weeks, or even months before an impending failure happens, allowing users time to take remedial action before a fault becomes permanent and before a failure ever occurs.
Unfortunately, conventional predictive and preventive maintenance approaches fail to scale down into advanced devices to meet the safety, performance, and cost challenges faced by vehicle manufacturers, OEMs, and fleet managers. In response, proteanTecs developed solutions that not only enable predictive and preventive maintenance but transform automotive electronics into intelligent system sensors: Continuous Performance Monitoring, Mission Profile Monitoring, and Degradation Monitoring. By combining deep data generated by on-chip agents with machine learning algorithms, the company provides advanced and predictive analytics during the car’s operational lifetime. Automotive SoC manufacturers have turned to proteanTecs to enhance their chips with these in-situ monitoring applications. In addition, leading automotive OEMs are pushing down these requirements into their supply chain as they move toward the goal of zero failures.
The proteanTecs Continuous Performance Monitoring application delivers in-vehicle real-time monitoring of the system’s unique performance indicators via embedded on-board software. Changing environmental and operational conditions, dynamic mission profiles, latent production defects, application stress, and device aging lead to fluctuations and deviations from the expected performance and reliability envelopes. By monitoring variations caused by a malfunction or unexpected behavior of electronic components, customers can reduce reliability risks, violations of operational requirements, or compromises to the vehicle’s functional safety.
The Continuous Performance Monitoring application measures operational health in mission-mode.
The Degradation Monitoring application monitors parametric performance values extracted from the proteanTecs on-chip agent data, which allows it to forecast the system’s incoming anomalous behavior and predict its remaining useful life. This application detects and measures impact of wear-out and degradation: it determines whether critical thresholds are exceeded, randomly or systematically, and alerts on faults before failures. These forecasts can trigger appropriate services and actions to avoid unwanted vehicle downtime.
The Degradation Monitoring application measures remaining performance margins and predicts time-to-failure.
proteanTecs monitoring agents have received ISO26262 Automotive Safety Integrity Level B (ASIL-B) certification, reflecting our ongoing commitment to supporting mission-critical systems at the highest standards.
The solutions are part of the next wave of industry standardization, driving Predictive Maintenance compliance as part of ISO 26262 and Lane Monitoring for PHY Compliance as part of the UCIe (Universal Chiplet Interconnect Express).
In the 2023 AutoTech Breakthrough Awards, proteanTecs’ solutions won “Overall Connected Vehicle Innovation of the Year,” further demonstrating the growing importance of this type of monitoring technology.
The benefits of predictive and preventive maintenance in automotive electronics include:
Download our white paper to learn more about “Predictive and Prescriptive Maintenance in the Context of Automotive Functional Safety”.
Sources
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Leave a Reply