In less than 40 years since the inception of the first self-driving car, we have witnessed significant progress and changes in the automotive industry. However, the journey toward the widespread adoption of fully autonomous driving still has its last mile to cover. While the industry focuses on commercially viable, limited advanced driver-assistance systems, the ultimate goal is to achieve truly hands-free, fully autonomous driving.
As someone currently immersed in the self-driving car industry and constantly exposed to the rapidly changing landscape of emerging technologies, I aim to provide an overview of the current technological status in the autonomous driving domain. Additionally, I'll highlight which automotive manufacturers and software companies are leading the way, offering readers a preliminary understanding of our proximity to fully autonomous driving.
Many companies will initially pursue L2 and L3-level autonomous driving in the next five to ten years, which involves lower technical requirements. This approach allows them to integrate these technologies into consumer vehicles, comply with safety regulations, and gradually introduce new technologies to consumers.
Current Stage: As of November 2023, we mostly find ourselves still in the L2 level (partial automation) stage. Surprisingly, despite the term "autonomous driving" existing for a considerable time, there are no truly autonomous vehicles available in the market. Testing on roads requires human monitoring and intervention when necessary. Even the recent unmanned taxi services in Arizona, China, and downtown San Francisco operate within confined areas. Due to these limitations, they fall under the L4 level (high automation), with the final mile to reach L5 level full automation still ahead. While the goal hasn't been achieved, progress has been made, putting us on the right track to a technology that will benefit humanity.
Current Directions in the Self-Driving Car Industry:
Self-Driving Taxis: Companies involved in developing self-driving taxis include Waymo, AutoX, Cruise, DiDi, Pony.ai, Zoox, Aurora, Motional, and Optimus Ride (acquired by Magna). Notably, Cruise and Waymo have faced challenges in urban deployment, causing disruptions and leading to government orders to reduce the number of self-driving cars on the road.
Consumer Vehicles: Most vehicles on the market now come with driver-assistance technologies such as adaptive cruise control, automatic emergency braking, lane departure warning, and dynamic stability control. Leading companies in this sector are Tesla, XPeng Motors, and Huawei's autonomous driving division. Tesla, having sold millions of vehicles globally, offers a Full Self-Driving (FSD) subscription that assists with driving even in city centers.
Autonomous Trucks: Autonomous trucks at L4 or L5 levels can be particularly useful for long-haul routes on highways. Companies leading in this sector include Aurora, TuSimple, Embark Trucks, and Daimler Trucks.
What's Next? Technological advancements are rapid, with companies claiming to achieve full autonomy in certain years. However, these goals often face delays or are not met as expected. Autonomous driving is an intricate problem that requires technological prowess, extensive testing time, and manpower to collect vast amounts of data and repeated systems refinements. Engineers worldwide are collaborating to address this colossal challenge. While an exact timeline for the birth of fully autonomous driving cannot be determined, it's my subjective belief that we are getting closer to this ultimate goal, and we may unknowingly usher in this new era within the next decade.
Key developments in the self-driving car industry from the 1980s to March 2023:
1980s-1990s | - 1985: DARPA builds the first autonomous land vehicle. |
- 1987: Ernst Dickmanns created the VaMoRs Mercedes-Benz, equipped with cameras for highway driving. | |
- 1995: Navlab 5 by Carnegie Mellon University completes a 3000-mile cross-country journey. | |
2000s | - 2004: First DARPA Challenge, no successful completion. |
- 2005: Second DARPA Challenge, Stanford University takes first place. | |
- 2007: DARPA Urban Challenge focuses on simulated city driving. | |
- 2009: Google initiates the Google Car project. | |
2010s | - 2013: Four U.S. states approve self-driving car testing. |
- 2015: Tesla introduces automatic driving software update for Model S. | |
- 2016: First Tesla self-driving car crash in Florida. | |
- 2018: Uber's self-driving car involved in a fatal pedestrian accident in Arizona. | |
2020s-Present | - 2020: Industry consolidation; the rise of the Chinese market with Huawei and XPeng surpassing Tesla. |
- 2020-2022: Mergers and acquisitions, e.g., Aurora and Uber ATG, Amazon and Zoox, Toyota and Lyft Level 5, Cruise and Voyage. | |
- 2022: Argo AI, jointly invested by Ford and VW, announces closure. Companies shift focus to L2+ and L3 automation. | |
- 2023: Ouster and Velodyne merge; Ouster focuses on intelligent infrastructure. | |
- March 2023: XPeng launches XNGP intelligent driving system; Huawei unveils ADS 2.0, both not relying on high-precision maps. |
Credit: Information provided by the author, table compiled by DIGITIMES Asia, Nov 2023
About the author:
Eva Mo holds a position as a Senior Software Engineer in Self-Driving Cars at NVIDIA, a prominent American technology company. Their role likely involves contributing to the development of software systems for autonomous vehicles, leveraging NVIDIA's expertise in artificial intelligence and graphics processing units.
Mo holds a Bachelor's Degree in Electrical Engineering from the National Taiwan University (NTU) and a Master's Degree in Robotics from Carnegie Mellon University (CMU) in the United States. She is dedicated to promoting knowledge and awareness in self-driving cars and robotics industries.
