The Rapid Evolution and Expansion of ADAS Technologies
Advanced driver assistance systems (ADAS) are technology groups that enhance driving safety. They can include early warning systems and automation. By alerting drivers to potential hazards and employing self-activating systems, they lower the risk of collision. Over 70% of new cars in the UK feature autonomous emergency braking systems (AEB); this figure is higher in the US. Automakers are united in their commitment to expanding the use of ADAS systems in the future, as there is no doubting its significance in reducing crashes and consequent injuries. In this article, we trace the origins of ADAS and overview the early technologies employed by the systems. We describe its evolution to the present day and anticipate future trends.
Origins
The introduction of anti-lock brakes is the first step in ADAS evolution. The system, now ubiquitous, helped drivers avoid collisions by preventing the wheels from locking and avoiding skids – which effectively signal a loss of control of the vehicle. The system uses wheel rotation sensors to determine whether a particular wheel is decelerating rapidly (i.e. about to lock) and valves in the hydraulic brake lines to isolate the affected wheel from the braking operation. A pump is employed to restore some braking pressure to that wheel once it is in rotation (accelerating) again. A computer controls the system, analysing the wheel sensors’ signal and operating the valves. The system has evolved to include steering angle sensors and front-to-back brake biasing, forming the electronic stability control (ESC) systems we encounter today.
Core concepts of ADAS
The concepts at the core of ADAS systems are that they should enhance driver and road safety. Cruise control, in its non-adaptive form, has been featured in cars for many years. However, this basic form does not contribute to driver safety. It is an autonomous feature but not a safety feature. The adaptive cruise control (ACC) systems we encounter today can react to a change in pace, maintain a safe distance from the car in front, and stop if necessary. ADAS uses real-time data from numerous sensors, combined and processed, to give a meaningful view of objects surrounding the car. The process, known as sensor fusion, incorporates a large amount of data and relies on fast, accurate, fault-tolerant in-vehicle networks (IVNs) to provide the necessary communications integrity.
Data sources
- Automotive imaging: Automotive imaging can augment the driver’s view of their surroundings, eliminating blind spots. Additional cameras for monitoring occupants, including the driver, and gesture recognition, are likely future features.
- LiDAR: Light detection and ranging (LiDAR) sensors transmit a pulsed laser signal and obtain proximity information by measuring the time it takes for this signal to bounce off any encountered objects and return to the sensor. Initially used in cartography, automotive LiDAR has evolved to provide an accurate 3D perception of the environment in which the vehicle is placed. Particularly useful in vehicle-to-infrastructure (V2I) communication
- Radar: Radio detection and ranging (radar) works similarly, transmitting and receiving electromagnetic waves instead of light. It is beneficial in an automotive context, as it can sense the proximity and trajectory of other vehicles; the waves bounce off metallic objects. This is particularly useful in vehicle-to-vehicle (V2V) communication.
- Computer Vision: Computer vision refers to using video signals to detect and track objects or events to perceive a real-world scene, particularly useful in lane departure warning systems.
Effective collision avoidance systems rely on perceiving the proximity of other vehicles and objects in the surrounding environment. Vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication are essential elements. Collectively known as vehicle-to-everything (V2X), the systems use wireless local area networks (WLAN) or cellular communication to form an ad-hoc network with anything within its range. The system transmits and receives low data volume cooperative awareness CAM or basic safety messages BSM.
Levels of ADAS
A classification system exists that quantifies the automation associated with each level.