ADAS service diagnostics and calibration — Ensuring precision and safety

Beginning with the 2018 model year, advanced driver assistance systems (ADAS) have undergone a significant technological step change. These systems continue to evolve and are integral to modern vehicle safety and performance, transforming how vehicles interact with their environments. These systems, possessing features such as lane keeping, adaptive cruise control, and automatic emergency braking, rely on complex on-board perception systems to function effectively. It is my opinion that professional service personnel today need to possess a strong understanding of these systems and how calibration and diagnostics are critical to maintaining vehicle safety and performance. My goal with this article is to uncover the hows and whys around ADAS and the importance of recognizing the fact that your clients rely heavily on your skills to ensure that these systems are able to deliver upon their intended purposes.

The critical role of ADAS in vehicle safety and performance

ADAS technologies enhance driver safety by severely reducing human error, which accounts for a significant percentage of today’s accidents involving vulnerable road users (VRUs). According to the National Highway Traffic Safety Administration, features found in SAE J3016 Level 1 vehicles, like forward collision warning and lane-keeping assist, can reduce crash rates by up to 20 percent. These systems rely on precise data from the on-board perception systems to interpret the vehicle’s surroundings and respond appropriately. However, inaccurate calibration, occluded viewports, or service oversights can compromise functionality, leading to reduced performance or safety risks. For service professionals, ensuring that ADAS components are properly mounted, calibrated, and ready is critical for the vehicle to deliver performance within its operational design domain.

Understanding on-board perception systems

Perception systems are the backbone of ADAS, enabling vehicles to “see” and “understand” their environment just like we as humans do. These systems integrate multiple sensor types, each with unique strengths, to provide comprehensive situational awareness. Be sure to reference the two articles I wrote about understanding ADAS sensors in the May 2023 and September 2023 PTEN issues for more information.

Camera

Cameras are the most common ADAS sensors, capturing visual data for lane detection, traffic sign recognition, and pedestrian identification. High-resolution cameras process color and texture information, making them ideal for tasks requiring detailed visual interpretation. Today’s camera systems can often outperform human vision through various filtering and image interpretation techniques. However, their environment may lead to a short life and or occluded vision. The windshield is a harsh environment where temperature, interior off-gassing, and condensation can negatively affect camera performance. Today’s camera technological capabilities have resulted in some vehicle manufacturers removing sensors such as radar and ultrasonic to achieve either the same or better performance.

Radar

Radar sensors use radio waves to detect objects, measure distance, and assess differential speed (Doppler), and deliver valuable information in all-weather conditions. Forward-facing radar sensors typically operate in the upper 70-80 GHz range to avoid attenuation within our ground-level atmosphere.

Computer vision basics in ADAS

Computer vision is the cornerstone of ADAS perception, enabling vehicles to interpret sensor data and make real-time decisions. It encompasses several key processes, each critical to system functionality:

Object detection and classification

Object detection identifies entities like vehicles, pedestrians, or road signs, while classification categorizes them (e.g., “car” vs. “motorcycle”). Algorithms like YOLO (You Only Look Once) or other open-source detectors are commonly used. Cameras must be properly mounted, have a clear, unrestricted view, and be properly calibrated to the vehicle. Calibration ensures that the sensors can deliver information relative to the vehicle configuration, such as critical geometric positions within the vehicle like trim height, pitch, roll, yaw, and other factors. Accurate calibration ensures sensors provide reliable data for these algorithms, as misalignment can lead to poor performance and safety risk.

Semantic segmentation

Semantic segmentation divides an image into meaningful regions, labeling each pixel (e.g., “road,” “sidewalk,” or “vehicle”). This process is vital for understanding complex scenes, such as distinguishing drivable areas from obstacles. Calibration errors can distort segmentation, leading to unsafe driving decisions.

Pose detection

Pose detection estimates the position and orientation of humans and other living beings to understand and predict their movements. For example, basic open-source pose detection programs can detect a human's left/right eyes, ears, shoulders, arms, elbows, and other extremities to help predict future trajectories. It has been my observation that many are unaware of these capabilities, as most folks assume that the cameras are only used for lane lines and vehicle detection. This is why it is extremely important to increase that awareness and ensure that these systems can deliver top performance.

Object and event detection and response

We, as human drivers, have learned to respond to various driving conditions and situations. Our attentiveness needs to be top-notch, and we also need to be equipped with the skills to react swiftly and accordingly. ADAS is essentially raising the bar by ultimately bringing super-human abilities to the table. This process involves detecting dynamic events, like a vehicle braking suddenly, changing lanes abruptly, red light runners, and other similar events by delivering appropriate responses, such as automatic braking, path planning, and other vehicle movements. Neural networks integrate sensor data to assess such scenarios and act within milliseconds to keep VRUs safe. Performing service operations within the service guidance provided with up-to-date service information is critical to maintaining ADAS performance. There is no room for maintenance-induced failures.

Calibration and diagnostics: Ensuring precision

ADAS calibration ensures that sensors and systems operate within the manufacturer's specifications. There are two primary types: static calibration, performed in a controlled environment using targets and alignment tools, and dynamic calibration, conducted on-road to fine-tune systems under real-world conditions. Both require specialized equipment, such as ADAS calibration frames, laser alignment tools, and diagnostic scanners. In some cases, the vehicle manufacturer may offer calibration options where the service technician can choose static or dynamic. There are other cases where both are required.

When it comes to diagnostics, the system approach isn’t much different for ADAS. Is the issue a component fault, communication, or power distribution? Technicians use diagnostic trouble codes to pinpoint malfunctions, such as sensor failures or temporary faults. Scan data can be used to surface information such as sensor disablements. These events could be triggered by temperature, a blockage, etc. Electrical basics apply during diagnostics. Does the sensor have everything it needs? Power, ground, network connectivity? A clear understanding of these basics will help you solve most issues that surface.

Common issues in ADAS-equipped vehicles

Although these systems are quite sophisticated, ADAS systems are prone to issues that challenge service professionals. Below are common problems and their implications:

Sensor misalignment

Physical impacts, such as minor collisions, pothole contact, etc., can shift sensors, leading to inaccurate data. For example, a misaligned camera may fail to detect lane markings at the right vehicle position, disabling lane-keeping assist operations. The radar sensor can be subject to harsh conditions where mounting points can allow the sensor to pop loose. I have personally seen this happen several times on Honda vehicles.

Improper calibration can occur due to numerous reasons. Improper calibration space, mismeasurements, target sizes, and other factors can lead to improper calibration and degraded system performance. The service technician shall not produce maintenance-induced errors.

Best practices for service professionals

To address these challenges, technicians should adopt the following practices:

1. Invest in training: Stay updated on ADAS technologies through training and certifications. 

2. Use advanced tools: Workshops should be equipped with the proper tooling to get the job done within the manufacturer's allowed tolerances. The closer you can get your targeting system to zero, the better the outcome. 

3. Follow protocols: Adhere to manufacturer-specific calibration and diagnostic procedures to ensure compliance and safety.