China claims breakthroughs in autonomous vehicles
Autonomous vehicles represent an estimated $2 trillion global market, and are expected to triple in value over the next decade. They are not just the future of transportation, but also have massive applications for the intelligent systems at the center of both U.S. and Chinese military plans.
In this economic and military race for the future, the Chinese Academy of Sciences’ Shanghai Institute of Microsystems and Information Technology (SIMIT) claims to have made two new breakthroughs toward overcoming one of the most crucial speedbumps: autonomous vehicles’ ability to perceive their surrounding environment and track movements. Their gains in the “brain” of autonomous vehicles are significant, as they come amidst other fast-paced developments with the “eyes” of vehicles, and signal substantial progress in China’s quest for autonomous vehicles.
SIMIT’s two breakthroughs relate to how autonomous vehicles visually perceive their surroundings and understand objects’ movements within them. When mapping out terrain for navigation, autonomous vehicles typically flatten the three-dimensional space into a two-dimensional grid projection. Although this method is quite effective for reducing calculation time, it introduces distortions and misrepresentations of the environment. SIMIT’s first breakthrough was to improve the method for streamlining three-dimensional representations, making them a more efficient alternative to two-dimensional projection.
The second advance refines this method by improving the vehicle’s ability to track an object as it moves across this three-dimensional representation and to predict its future location based on current trajectory. Think of this as akin to how the brain of a human naturally projects the flight path of a ball in the air or a bicycle rolling down the street.
These reported breakthroughs come amid intense competition in the fields of autonomous vehicles, advanced driver-assistance systems, and unmanned aerial vehicles. Indeed, any progress towards autonomous capabilities inherently has military applications. For example, commercial drones have been used for reconnaissance and as munitions on Ukrainian battlefields, while the People’s Liberation Army, the Pentagon, and many other militaries around the world are developing autonomous vehicles. Under China’s policy of Military-Civil Fusion, SIMIT’s improved ability to map and track targets thus has value to not just associated car manufacturers, but also the PLA.
China and the PLA have been doggedly pursuing improvements and new capabilities in this field for years. These particular breakthroughs from SIMIT were realized through contributions from China’s National Natural Science Foundation, the Chinese Academy of Sciences, and the “Science and Technology Innovation 2030” project, which directs funding towards organizations working towards applications for artificial intelligence, such as advanced drone technology. Similarly, the Natural Science Foundation focuses on funding high-tech priority areas designated by the Chinese government, including “intelligent unmanned system technology” such as that developed by SIMIT.
Beyond these developments in the “brain” of autonomous vehicles, China is also pushing ahead with the “eyes” that send the signals to the computer to interpret. At the front line of this race towards fully-realized autonomous vehicles is the push to dominate the market for LiDAR (light detection and ranging) systems. Akin to how radar works with radio waves, LiDAR uses lasers to mark distances and create three-dimensional maps of the environment in real-time. Initially intended for landscape and oceanographic mapping, LiDAR quickly became the main “eyes” for self-driving vehicles and driver assistance systems in both military and civilian vehicles. Without LiDAR, any improvements in how autonomous vehicles interpret real-world data would be useless.
U.S. companies controlled the LiDAR market until relatively recently, when Chinese companies rose seemingly out of nowhere to dethrone U.S. dominance through state subsidies and other support. In particular, PRC documents from 2020 note that numerous government organs were directed to surge their aid for “intelligent vehicles” and “establishing an integrated sensory system” for urban environments.
This growing Chinese presence was first spotted in August of 2023, when a Congressional Research Service report highlighted the Chinese push into the LiDAR market. The report found that on top of the typical subsidies provided by Chinese industrial policy, the Chinese government is also allocating resources to acquire foreign IP, encourage foreign manufacturers to establish factories in China, and push for the promotion of Chinese standards on the market. This has led to numerous concerns about how PRC automobile sensors like cameras and LiDAR gather massive amounts of data about their surroundings and even passengers, which could easily be used to gather intelligence.
Following these revelations, the House Committee on the Strategic Competition Between the United States and the Chinese Communist Party asked the U.S. Government to investigate PRC LiDAR firms, their relationship with the PRC government, and the possibility of PRC LiDAR products entering U.S. Government supply chains.
In January, the U.S. Defense Department added PRC LiDAR manufacturer Hesai to the List of Chinese Military Companies Operating Directly or Indirectly in the United States (although it may remove the company from the List pending a court decision). Finally, in July of 2024 the U.S. House Committee on Commerce, Justice, Science, and Related Agencies recommended the U.S. Department of Commerce Bureau of Industry and Security investigate PRC LiDAR manufacturers for ties with the Chinese Communist Party (CCP) and contributions to the PLA. Lawmakers also urged the Department Commerce to add such companies to the department’s Entity List, which would make it difficult for the Chinese companies to expand within the United States.
These concerns around PRC LiDAR manufacturers are well-founded. Despite Hesai’s convoluted corporate structure, multiple Virgin Islands shells, and U.S. IPO, the company was found to have connections to the China Electronics Technology Group Corporation (CETC), a company founded to supply advanced electronics to the PLA. RoboSense, another China LiDAR manufacturer, has demonstrated ties to the Harbin Institute of Technology (HIT), itself one of the PRC’s “Seven Sons of National Defense” universities. RoboSense’s CTO is a former professor, while the school’s CCP Committee Secretary visited the company headquarters in 2023 and promoted HIT as a feeder for the company’s technical needs. Hesai and RoboSense have both made substantial gains in the LiDAR market, together growing sales from around 10,000 units in 2020 to nearly 500,000 last year.
Such expansion points to China’s growing weight in the world of autonomous vehicles, with the new advances in their suite of sensors set to accelerate that trend. If Chinese companies succeed in solidifying their dominance of the autonomous vehicle and LiDAR markets, the security implications are profound. There is already great concern over modern Chinese-manufactured automobiles’ ability to sense their surroundings with driver-assistance systems such as rear-view cameras and blindspot warning systems, and their ability to collect far more personal data than consumers realize. Although the U.S. Government is moving to counter these concerns, it is still in the information gathering phase and has not yet implemented controls to address these risks.
As the importance of autonomous vehicles, their suite of sensors, and the security implications therein grows, so too will the need to maintain U.S alternatives to adversarial advancements. If the Chinese Academy of Sciences’ SIMIT continues to refine their capabilities, this need will only rise in tandem.
Thomas Corbett is a research analyst with BluePath Labs. His areas of focus include Chinese foreign relations, emerging technology, and Indo-Pacific security studies.
P.W. Singer is a best-selling author of such books on war and technology as Wired for War, Ghost Fleet, and Burn-In; senior fellow at New America; and co-founder of Useful Fiction, a strategic narratives company.
Comments are closed.