Assisted Driving

Assisted driving, generally referring to Advanced Driver Assistance Systems (ADAS), is a suite of technologies that utilises onboard sensors to monitor the surrounding environment in real time, helping drivers operate vehicles more safely and comfortably.

Its primary role is to assist human drivers rather than fully replace them, aiming to mitigate safety risks stemming from human error, and serving as a vital foundation towards higher levels of autonomous driving.

🧠 Working Principle: Perception-Decision-Execution

Operating like a vigilant co-driver, the assisted driving system functions through a closed loop of "Perception-Decision-Execution":

1. Perception Layer (Eyes and Ears): Scans the surrounding environment clearly via multiple complementary sensors.

   • Cameras: Responsible for identifying lane markings, road signs, traffic light signals, as well as detecting pedestrians and other vehicles.

   • Millimetre-wave Radars: All-weather sentinels, adept at accurately measuring the distance and speed of objects ahead in poor visibility conditions such as heavy rain, fog, or darkness.

   • LiDAR: Generates high-precision 3D point cloud maps to accurately detect obstacle contours and positions at long ranges, further bolstering perception capabilities.

   • Ultrasonic Sensors: Dedicated primarily to short-range proximity detection, serving as the core hardware for automated parking features.

   The current industry trend centres on multi-sensor fusion, allowing different sensors to complement one another's strengths, eliminating single-sensor blind spots, and delivering comprehensive situational awareness.

2. Decision Layer (Brain): High-performance computing platforms (domain controllers) process and fuse sensor data, executing sophisticated algorithms (such as end-to-end models) to evaluate driving conditions and make real-time decisions.

3. Execution Layer (Hands and Feet): Translates commands from the decision layer into vehicle control, managing steering, throttle, and braking inputs to complete specific driving manoeuvres.

📜 Core Functions: From Warnings to Active Intervention

Assisted driving features span multiple levels, ranging from driver alerts to active vehicle control:

• Warning Systems: Alerting drivers to impending frontal impact risks (FCW), monitoring approaching vehicles in blind spots (BSD), or warning when the vehicle inadvertently drifts out of its lane (LDW).

• Active Intervention: Autonomous Emergency Braking (AEB) — vehicles fitted with AEB can reduce rear-end collisions by over 35%; highway driving assistance managing speed and lane positioning (ACC and LKA); and automated parking assistance (APA).

📊 Classification and Industry Landscape

Based on the Society of Automotive Engineers (SAE) standards, mainstream mass-production systems on the market today predominantly cover Level 1 (Driver Assistance) and Level 2 (Partial Driving Automation).

At Level 2, the system can concurrently control both lateral (steering) and longitudinal (acceleration and deceleration) motion; however, the driver must remain fully attentive to road conditions and be prepared to take over at any moment. Higher-level Navigation on Autopilot (NOA) capabilities are also gaining traction, enabling point-to-point assisted driving in dedicated scenarios such as highways.

With rapid technological advancements, leading automotive and tech firms have rolled out assisted driving systems that have served millions of motorists and clocked tens of billions of test and production kilometres. Powered by continuous algorithm iterations and computing upgrades (such as World Models and Vision-Language-Action (VLA) models), the driving experience is progressively emulating the natural, intuitive instincts of a seasoned driver.