Despite over a decade of technological advancements and testing in cities like Phoenix and San Francisco, robotaxis still faces significant hurdles.
Companies like Waymo, Cruise, and Amazon’s Zoox are actively developing the concept of commuting via robotaxi without the need for car ownership. These companies have spent over a decade advancing autonomous vehicle technology, and testing robotaxi services in cities such as Phoenix and San Francisco. Recently, Waymo, a subsidiary of Alphabet, expanded its service to parts of Los Angeles, operating without a safety driver. Despite advancements, the rollout of robotaxis has faced challenges. Some vehicles have experienced issues, stopping in the middle of roads, making illegal turns, and causing accidents. Additionally, self-driving cars still struggle in adverse weather conditions like snow and rain, which impair their sensors.
Tesla is also entering the robotaxi market, with plans to reveal its own model. However, widespread adoption remains a distant goal. An assistant professor at Northeastern University, believes the technology is not yet ready for mainstream use. He points out that autonomous vehicles are still highly specialized, relying on lidar sensors, GPS, and cameras to navigate. He explains that self-driving cars must make numerous decisions per second to understand their environment, a process that is not yet perfect. The vehicles use lidar to map their surroundings, but interpreting this data to distinguish safe from unsafe conditions remains a challenge. For instance, a car must decide whether an object on the road is a pothole or a person, a critical safety decision.
The broader industry progress is hard to assess due to companies’ incentives to overstate their technology’s capabilities. Tesla is currently facing a class-action lawsuit over claims of misleading users about its self-driving technology. The National Highway Traffic Safety Administration has also opened investigations into Waymo, Zoox, Tesla, Cruise, and Ford regarding their autonomous vehicles or advanced driver-assist systems.
The researcher emphasizes the need for transparency in deploying these technologies. Autonomous cars are tested on public streets, affecting pedestrians who have not agreed to be part of these experiments. Significant improvements in artificial intelligence are required for these technologies to advance. While the hardware is advanced, he believes innovation in software algorithms is crucial for solving current problems. Large engineering teams are working on aspects like planning, perception, and prediction to enhance the technology’s capabilities.
