The Popularization Process of Advanced Driver Assistance and Autonomous Driving Applications

By Jeff Lee 2023-04-10

ADAS (Advanced Driver Assistance Systems) has become a prominent technological competition among car manufacturers since 6-7 years ago. Since then, autonomous driving has become an irreversible global trend. At this point, various technology startups have sprung up like mushrooms after the rain, and their competition with car manufacturers has begun to develop in an overlapping manner, reaching a critical moment of L2+ popularization. Next, under the increasingly strict evaluation of safety assessment organizations, L3 will face practical challenges, while L4, although technically difficult, once its technology is mature, will become a milestone that subverts our mobility modes. This point in time is expected to arrive around 2030.

According to a recent report by the well-known research institution Statista on L2 and above autonomous driving technology applications, nearly two-thirds of passenger cars sold globally in 2025 will have advanced assistance driving or even autonomous driving functions. About 60% of them will be L2 or L2+, and about 4% will have L3 highway pilot capability. With the popularization of these intelligent driving functions, the incidence of traffic accidents will gradually decline year after year (although there are still some accidents caused by excessive reliance on assisted driving). Of course, in terms of the application timeline for higher-level, fully autonomous driving systems (L4/L5) technology, Statista estimates that only about 3% of new cars will have L4 functionality in 2030; that is to say, at that time, fully autonomous driving systems will have just begun to hit the road. In 7-8 years, L2 and L2+ will be the mainstream of intelligent driving systems. Companies that lack financial resources or can’t capitalize on local scenes will gradually become obsolete in the competition for the development of fully autonomous driving technology. At the same time, global automotive safety-related authoritative organizations such as Euro NCAP (European New Car Assessment Program) will inevitably begin to tighten their scoring standards. The version of Euro NCAP was revised in 2020 and divided into four aspects: adult occupants, child occupants, vulnerable road users, and safety assistance. Among them, "safety assistance" is the evaluation closely related to assisted driving and autonomous driving systems:

  • AEB (Automatic Emergency Braking) is evaluated in several scenarios.
  • DMS (Driver Monitoring System).
  • Speed limit reminder.
  • LKA (Lane Keeping Assist) and ELK (Emergency Lane Keeping).
  • LDW (Lane Departure Warning) and BSD (Blind Spot Detection).

Given that these safety features are already well-known proprietary terms in the current market, it's all thanks to the constant updating of evaluation criteria by these safety assessment organizations. With wider adoptions for L2+ and even L3, what direction will Euro NCAP take in its next steps?

  • AES (Automatic Emergency Steering) technology will be one of the key focuses of global car makers' research and development, and the evaluation methods will be more explicit.
  • AEB (Automatic Emergency Braking) optimization will address more complex traffic scenarios such as intersections.
  • V2X (Vehicle-to-Everything) standards will promote "vehicle-to-cloud" collaboration for a safer and smoother driving environment. For example, vehicles will need to have the ability to send and receive messages such as "emergency electronic braking lights," "motorcycles approaching," or "road construction ahead." This also means that the deployment of 5G and beyond (6G era in 2030) for transportation infrastructure is an urgent task for governments worldwide.
  • Grading of autonomous driving systems’ intelligence levels. I believe Euro NCAP's classification and definition will be more accurate and understandable to consumers than the US SAE classification (Level 0~5). This will include clarifying the usability, identifying functions clearly, and even requiring car manufacturers and consumers to establish relevant agreements. Euro NCAP may define several scenarios, such as parking, urban driving, interurban driving, traffic jam, and highway driving.
  • The ability to recognize traffic signs, such as "one-way" and "no-entry." Because different countries have different traffic signs, this function’s localization will require further time for development.
  • Introducing virtual testing to simulate more different traffic scenarios.

In the next few years, as technology companies and major car manufacturers continue to innovate and release new autonomous driving systems, safety evaluation organizations will also set more stringent and precise evaluation standards. Taiwan’s T-NCAP began operating last year under the auspices of VSCC. How will this third-party organization monitor the Taiwanese consumer market to embrace the new world of electric and smart vehicles? In addition to past practices of using EU regulations as a basis for formulating Taiwan's emissions and safety regulations, T-NCAP should collaborate with ARTC and the academic and research community, such as ITRI, to develop a safety evaluation system that truly meets the needs of Taiwan's traffic environment and chart its path forward.