1 hour ago
3
ARTICLE AD BOX
Waymo is using DeepMind's Genie 3 artificial intelligence (AI) model to create realistic digital environments where its autonomous vehicles can train on rare and difficult driving scenarios.
The Alphabet-owned self-driving company recently published a blog post announcing its new Waymo World Model. This AI model is built on Genie 3's ability to construct virtual environments from text prompts. The system generates synthetic driving footage and depth perception data, as if captured by vehicle-mounted cameras and lidar sensors.In a statement to Bloomberg, a Waymo spokesperson said, “Traditional AV simulation models are constrained by the on-road data they collect,” and the new world model “allows us to explore situations that were never directly observed by our fleet.”Waymo's system can also convert real-life dashcam datasets into scenes and depth maps for vehicle simulations. The company said this combination will make its autonomous vehicle systems more reliable in uncommon scenarios and help expand Waymo's self-driving services across more markets."This will enhance Waymo's ability to safely scale our service across more places and new driving environments," the spokeswoman added.
Last week, Google DeepMind’s Genie 3 model gained major attention with demonstrations of its world-building capabilities. This also triggered a sell-off in companies that provide game development and graphics creation tools.
How Google DeepMind's Genie 3 model can help Waymo
By incorporating Google's Genie 3 model, Waymo could gain momentum for its plans to expand to about a dozen cities this year. Simulation is just one tool it uses to prepare its autonomous systems for specific situations and validate their safety, the spokesperson noted. In December 2025, the company said there is "no substitute" for its real-world driving experience. Recently, Alphabet reported Waymo had surpassed 20 million autonomous trips that month.However, the company spokesperson declined to comment on whether it has conducted simulations of stopped school bus encounters or mass power outages, such as the one that disrupted its San Francisco operations last year. Yet, she noted the Waymo World Model "can simulate virtually any scene."Waymo is currently facing safety probes from US authorities after a series of software mishaps in recent months. The National Highway Traffic Safety Administration and the National Transportation Safety Board are investigating multiple incidents in which Waymo failed to stop for parked school buses in Austin. These violations have also prompted the company to issue a voluntary recall of its software.Beyond Waymo, other robotaxi operators and AI companies have been seeking additional data sources to improve their models.
Nvidia, which supplies chips and AI models for developers of self-driving technology, has partnered with ride-hailing company Uber to collect millions of hours of robotaxi-specific driving data to fuel driverless model training and validation.SoftBank Group-backed Wayve, which plans to trial robotaxis on the Uber platform in the UK this year, has announced its own world model to generate synthetic driving data. Elon Musk’s Tesla has also said that it has built a similar simulator. Bigger training data sets will be crucial for Waymo and help avoid such incidents.
English (US) ·