inDrive Partners with Pony.ai and Nvidia for Robotaxi Tests in Central Asia

When you think of autonomous vehicle testing grounds, the sprawling, snow-drenched expanses of Central Asia rarely top the list. Yet ride-hailing app inDrive is preparing to shift that geographic baseline. The company, known for letting passengers and drivers negotiate fares directly, has signed a memorandum of understanding to bring self-driving technology to Kazakhstan's capital, Astana, by the first half of 2027.

Rather than building proprietary autonomy stacks from scratch, inDrive is leaning on an external hardware and software supply chain. Under the proposed framework, Chinese autonomous driving firm Pony.ai will supply its Gen-7 robotaxis to run on inDrive's platform, while Nvidia provides the underlying artificial intelligence infrastructure, including its Hyperion platform and Halos safety system. Initially, the trials will feature a safety operator behind the wheel, with driverless commercial scaling set as the ultimate objective.

inDrive plans to roll out Pony.ai's Gen-7 autonomous vehicles integrated with Nvidia hardware during trials in Astana.

The Capital-Light Playbook for Emerging Markets

For inDrive, entering the robotaxi race requires a different financial calculus than the billions spent by dedicated autonomy shops in the United States and China. According to inDrive CTO Yuri Misnik, the company's operational design remains capital-light. By partnering with established AV developers and hardware providers rather than incubating neural networks internally, the ride-hailing platform hopes to bypass early-stage research costs while expanding into regional markets.

Nvidia's hardware role is central to keeping these deployment costs down. Ali Kani, Nvidia's vice president of automotive, noted that the Hyperion platform offers companies a ready-made foundation of sensors and computing silicon. Instead of custom-engineering sensor suites and sensor-fusion algorithms, manufacturers can utilize pre-integrated systems alongside Nvidia's Alpamayo model and synthetic data generation tools to simulate edge cases across multiple driving environments.

Navigating Astana's Brutal Winter Conditions

Testing in Astana—a city of roughly 1.7 million residents characterized by months of heavy snow, freezing temperatures, and black ice—offers a severe proving ground for autonomous stacks traditionally trained in milder climates. inDrive founder and CEO Arsen Tomsky emphasized that evaluating autonomous systems against harsh winter weather will yield insights that translate directly to other regions facing similar meteorological hurdles.

The timeline aligns with local regulatory shifts. Last month, the government of Kazakhstan formally outlined regulations to permit autonomous vehicle testing on specific public roads, paving the way for pilot programs slated for 2027 across Astana and two other regional hubs. This legal framework matches inDrive's broader expansion strategy, which includes a separate agreement signed in December with Saudi Arabia's AI Driver to explore WeRide-powered autonomous vehicles in Riyadh.

As Pony.ai accelerates its global footprint—aiming to scale its fleet past 3,500 robotaxis across more than 20 cities worldwide—its partnership with inDrive signals how regional ride-hailing networks are securing access to advanced autonomy without waiting for Silicon Valley giants to export domestic models.