For International Release
The SHINDEV Research Team believes that a recent fatal autonomous driving accident in the United States has once again placed the safety and commercial viability of self-driving technology at the center of public and industry debate.
In the incident, a driverless taxi operated by Cruise, the autonomous driving subsidiary of General Motors, collided with a pedestrian. Although the vehicle performed an emergency stop, it subsequently initiated an automated pull-over maneuver, during which the pedestrian was subjected to secondary crushing and dragging, ultimately resulting in death.
This tragedy has not only triggered widespread public concern regarding the reliability of autonomous driving systems, but has also undermined long-standing industry confidence in the maturity and readiness of the technology.
After more than a decade of development, autonomous driving technologies have diverged into two primary pathways. One is the incremental approach, represented by Tesla, which evolves from driver assistance features on mass-produced vehicles toward full autonomy. The other is the leapfrog approach, exemplified by Cruise, which targets L4-level Robotaxi deployment directly.
The significance of this incident lies in the fact that Cruise is widely regarded as one of the most capable players in the global autonomous driving sector. Since being fully acquired by General Motors in 2016 for nearly USD 1 billion, Cruise has attracted major strategic investors including SoftBank Vision Fund, Honda, Microsoft, T. Rowe Price, and Walmart. Total funding has approached USD 10 billion, with valuation once exceeding USD 30 billion.
In August 2023, Cruise obtained regulatory approval to provide paid, round-the-clock autonomous taxi services, widely viewed as a milestone for commercialization. However, the accident was followed by the departure of key founding executives and General Motors’ announcement of a USD 10 billion share repurchase program, amplifying market concerns over the sector’s stability.
From a technical perspective, the incident was not caused by a failure in obstacle detection. The autonomous system successfully identified the obstacle and executed an emergency stop, indicating strong forward perception capability. However, after detecting an abnormal situation and executing an automated pull-over strategy, the system failed to detect anomalies beneath the vehicle chassis, leading to the fatal secondary injury.
SHINDEV Research notes that this highlights a critical issue: whether autonomous systems possess independent, high-priority safety redundancy mechanisms capable of overriding flawed system logic in exceptional scenarios.
Advanced autonomous driving relies heavily on brake-by-wire systems. Currently, Electro-Hydraulic Braking (EHB) is the mainstream solution, retaining hydraulic fallback mechanisms that ensure braking safety in case of system failure.
The next-generation Electro-Mechanical Braking (EMB) system eliminates hydraulic components entirely, offering superior integration, reduced complexity, and better compatibility with intelligent vehicle architectures. However, its high degree of electrification introduces challenges related to reliability, thermal management, regulatory compliance, and redundancy. As a result, large-scale commercialization remains unlikely in the near term.
The SHINDEV Research Team emphasizes that autonomous driving safety cannot be attributed to any single subsystem. The incident reveals limitations in current sensor coverage, particularly regarding underbody detection and post-stop environmental verification. Future high-level autonomous vehicles must incorporate multi-modal sensing and system-level redundancy from the design stage.
L4 autonomous driving depends heavily on external infrastructure. Roads of the future must integrate sensing, communication, and data platforms to enable real-time interaction between vehicles, infrastructure, and cloud systems, forming a comprehensive intelligent transportation ecosystem.
In the short term, SHINDEV Research recommends a cautious approach to autonomous driving commercialization, prioritizing safety redundancy and system coordination. Over the medium to long term, continued advances in artificial intelligence, brake-by-wire systems, and intelligent infrastructure are expected to drive the gradual maturation of autonomous driving, positioning it as a core pillar of future mobility.
