Ride-hailing platform inDrive plans to begin testing autonomous robotaxis in Astana, Kazakhstan, during the first half of 2027, working with Chinese autonomous-driving company Pony.ai and US semiconductor giant Nvidia. The proposed programme aims to explore how Pony.ai’s seventh-generation robotaxis could be integrated into the inDrive app, potentially allowing passengers to request autonomous rides through an established mobility platform.
The collaboration brings together three different parts of the autonomous mobility ecosystem: inDrive’s ride-hailing network, Pony.ai’s self-driving vehicle technology and Nvidia’s computing and safety infrastructure. Initial tests are expected to use safety operators, with the longer-term objective of assessing whether driverless services can be introduced. The plan is not yet a confirmed commercial launch, and its progress will depend on technical testing, regulatory approvals and operational readiness.
Key Takeaways
- Target date: Supervised robotaxi testing is planned for the first half of 2027.
- Test location: Astana, Kazakhstan, is the proposed initial deployment city.
- Partners: inDrive, Pony.ai and Nvidia are collaborating on the initiative.
- Vehicle technology: The programme will explore Pony.ai’s seventh-generation, or Gen-7, robotaxis.
- Nvidia’s role: Its DRIVE Hyperion platform and Halos safety system are expected to support the programme.
- Initial safety arrangements: Tests are expected to include a safety operator.
- Commercial status: The initiative is at the planning and testing stage, not a confirmed fully driverless passenger service.
What Is inDrive Planning With Pony.ai and Nvidia?
The companies have signed a memorandum of understanding to explore integrating Pony.ai’s Gen-7 robotaxis into the inDrive app and conducting supervised testing in Kazakhstan’s capital.
The agreement addresses a key challenge for autonomous mobility: connecting a self-driving vehicle to a service that customers can actually use. Developing a vehicle capable of navigating roads is only one part of operating a robotaxi business. Companies also need passenger booking systems, fleet management, maintenance, customer support and processes for handling incidents.
inDrive already operates a ride-hailing platform that connects passengers with drivers. By exploring robotaxi integration, the company could assess how autonomous vehicles fit into its existing booking and dispatch system.
Pony.ai would provide the autonomous-driving technology and vehicles, while Nvidia’s computing and safety platforms would support the technical architecture. The proposed collaboration could allow the partners to evaluate how the system performs under local traffic, weather and operating conditions.
However, the memorandum does not establish that a full commercial service will launch in 2027. The immediate objective is to test feasibility and performance.
What Each Company Brings to the Partnership
| Company | Primary role | Potential contribution |
|---|---|---|
| inDrive | Ride-hailing platform | Passenger booking, app integration and mobility operations |
| Pony.ai | Autonomous-driving technology | Gen-7 robotaxis and self-driving systems |
| Nvidia | Computing and safety technology | DRIVE Hyperion platform and Halos safety system |
The division of responsibilities reflects the way autonomous mobility is developing. Rather than building every part of the service internally, companies can combine vehicle technology, computing infrastructure and an existing passenger network.
For inDrive, the partnership provides a route to explore autonomous rides without having to develop its own self-driving vehicle from scratch. Pony.ai gains another potential overseas deployment opportunity, while Nvidia extends the reach of its automotive computing ecosystem.
The ultimate value of the arrangement will depend on whether the partners can move beyond technical demonstrations and establish a reliable, commercially viable service.
Why Astana Is an Important Testing Location
Astana offers the partners an opportunity to evaluate autonomous vehicles in a Central Asian environment, including cold weather and winter road conditions.
Autonomous-driving systems must interpret their surroundings using cameras, radar, lidar and other sensors, depending on the vehicle’s configuration. Snow, ice, reduced visibility and changing road conditions can make perception and planning more difficult.
Testing in different environments can help developers identify limitations that may not emerge in more favourable conditions. These tests can examine how vehicles detect road markings, recognise other road users, respond to obstacles and maintain safe behaviour when conditions change.
The location may also help the companies evaluate the operational requirements for autonomous mobility in a city outside the major markets where robotaxi services have developed most rapidly.
Nevertheless, the choice of Astana does not by itself establish that the technology is ready for winter operation without human supervision. Performance will need to be assessed through controlled tests, and any future service will need to meet local regulatory and safety requirements.
Nvidia’s DRIVE Hyperion and Halos Systems
Nvidia’s involvement is significant because autonomous vehicles require substantial onboard computing to process sensor information and make driving decisions in real time.
The DRIVE Hyperion platform is designed to provide a hardware and software foundation for vehicles intended to support advanced autonomous-driving functions. It brings together computing, sensor and system architecture components that automakers and autonomous-driving developers can use in building their systems.
Nvidia’s Halos safety system is intended to support safety engineering across autonomous vehicle development and deployment. The broader objective is to help developers design, validate and operate autonomous systems with safety requirements incorporated throughout the process.
Pony.ai has an established relationship with Nvidia. In April 2026, Pony.ai announced a new-generation autonomous-driving domain controller developed with Nvidia technology, including the DRIVE Hyperion platform and DRIVE AGX Thor computing. The collaboration was intended to support Level 4 autonomous-driving applications.
Level 4 automation generally means a vehicle can perform the driving task within a defined operational domain without requiring a human to drive it. This does not mean the vehicle can operate autonomously in every road, weather or geographic condition.
The Astana programme will need to demonstrate how the relevant technology performs under the specific conditions in which the partners intend to test it.
How Robotaxis Could Fit Into the inDrive App
Integrating robotaxis into a ride-hailing app could make autonomous mobility easier for customers to access. Instead of using a separate application, passengers could potentially request an autonomous ride through an interface they already know.
For the operator, an app-based model can also help coordinate vehicle availability, trip requests, passenger communications and fleet utilisation. These operational systems are important because autonomous vehicles need to be dispatched efficiently to make the economics of a service work.
But integrating a robotaxi into an app is more complicated than adding a new vehicle category. The platform must account for service areas, vehicle eligibility, pickup locations, passenger instructions, remote assistance and procedures for situations in which a vehicle cannot complete a journey.
The planned supervised testing phase should help the partners understand how the service works in practice. No final details have been announced about fares, fleet size, public access or the geographic coverage of the proposed pilot.
Pony.ai’s International Robotaxi Expansion
The agreement also fits into Pony.ai’s broader international expansion strategy. The company has been developing autonomous-driving systems and operating robotaxi services in China while working with partners to enter additional markets.
Pony.ai and its partners have announced projects in several international locations. In September 2026, Pony.ai and Verne began fully driverless robotaxi test rides in Zagreb, Croatia. The company has also reported progress on robotaxi operations in Doha, Qatar.
These projects illustrate the variety of deployment models available to autonomous-driving developers. Some involve supervised testing, while others have progressed to driverless rides within defined operating areas.
The inDrive partnership would provide another opportunity to evaluate Pony.ai’s technology outside China. It could also test how an autonomous vehicle operator and a ride-hailing platform divide responsibilities for customer service, fleet management and local operations.
However, each market has its own regulatory requirements, road conditions and operating economics. Progress in one city does not automatically guarantee that a service can be introduced in another.
The Business Challenge: Moving From Testing to Commercial Service
The robotaxi industry faces a significant gap between demonstrating autonomous driving and running a dependable commercial fleet. A service must operate safely, achieve sufficient vehicle utilisation and manage costs associated with vehicles, maintenance, computing, insurance and remote support.
A safety operator during early tests can help monitor the system and intervene when necessary. Moving to fully driverless operations would require the companies to demonstrate that their systems can meet applicable safety standards and operate within an authorised service area.
Regulatory approval will be particularly important. Authorities may require testing permits, vehicle certifications, insurance arrangements and operational safeguards before allowing passengers to use driverless vehicles on public roads.
The partners have not announced a confirmed commercial launch date, expected fleet size or financial investment for the Astana programme. Those details will be important in determining whether the initiative remains a limited pilot or develops into a larger mobility service.
The Bigger Picture
The proposed Astana programme reflects a wider shift in autonomous mobility, where technology developers are partnering with ride-hailing platforms and computing providers to bring robotaxis into new markets. The approach can reduce the need for a single company to build every part of the system, while allowing each partner to focus on its existing strengths.
For inDrive, the initiative could provide a foothold in autonomous ride-hailing without requiring it to develop its own self-driving technology. For Pony.ai and Nvidia, it offers another potential test environment and deployment pathway. The commercial significance will depend on the results of supervised testing, local approvals and the ability to operate vehicles safely and economically.
FAQs
1. When will inDrive begin robotaxi testing?
The partners are targeting the first half of 2027 for supervised testing in Astana, Kazakhstan.
2. Which companies are involved in the project?
The proposed programme brings together inDrive, Pony.ai and Nvidia. inDrive provides the ride-hailing platform, Pony.ai supplies the autonomous-driving technology, and Nvidia contributes its computing and safety systems.
3. Will the robotaxis be fully driverless from the start?
No. Initial tests are expected to include a safety operator. Fully driverless operations are a longer-term objective, subject to technical performance and regulatory approval.
4. What technology will Nvidia provide?
Nvidia is expected to contribute its DRIVE Hyperion platform and Halos safety system to support the autonomous-driving programme.
5. Will passengers be able to book robotaxis through the inDrive app?
The memorandum of understanding includes exploring integration of Pony.ai’s robotaxis into the inDrive app. Public availability, fares and the timing of any commercial service have not been confirmed.
Looking Ahead
The first major milestone will be whether the partners can begin supervised testing in Astana during the first half of 2027. The results will help establish how Pony.ai’s robotaxis perform in local traffic and weather conditions, and whether the inDrive platform can support the operational requirements of autonomous rides.
A commercial rollout would require additional progress on safety validation, regulation, fleet economics and customer experience. Until the partners disclose more details about the test programme and its results, the initiative should be viewed as an important planned trial rather than confirmation that driverless ride-hailing is about to become widely available in Central Asia.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
