DiDi R2 Robotaxi during fully driverless service trials ahead of planned UAE testing
The China-based trial is an important technical step. However, UAE deployment still depends on local validation, regulation and operating evidence.
| Key Distinction: R2 is now being trialled without a safety driver in designated Chinese areas. DiDi has not announced a live UAE R2 service. |
A China trial with clear UAE relevance
DiDi Autonomous Driving has started fully driverless service trials with its next-generation R2 Robotaxi in China.
The vehicle is operating in designated demonstration areas in Guangzhou and Beijing, according to the company’s announcement.
That detail matters. The trial is not taking place in Dubai or Abu Dhabi, despite the announcement’s strong UAE focus.
The UAE connection concerns DiDi’s next phase. The company identifies the country as its first international market and plans local testing.
For UAE readers, the R2 story is therefore about readiness, not arrival. A Chinese service trial cannot replace validation on local roads.
What DiDi has built into the R2
DiDi developed the R2 with GAC Aion as a purpose-built Level 4 Robotaxi, rather than converting a conventional passenger car.
Level 4 automation can handle the driving task without human intervention inside defined conditions and approved operating areas.
It does not mean the vehicle can drive everywhere, in every situation. Its operational limits remain central to any deployment decision.
DiDi says the R2 carries 33 sensors. These include LiDAR, cameras, 4D millimetre-wave radar, infrared cameras and audio sensors.
The company also claims more than 2,000 TOPS of GPU computing performance through a triple-domain central computing platform.
Those numbers sound impressive, but sensor counts alone do not establish safety. Integration, software behaviour and operational discipline matter more.
DiDi also says the vehicle platform meets five-star C-NCAP and Euro NCAP standards and includes layered system redundancy.
Independent readers should treat these specifications as manufacturer claims until supporting test records and market-specific approvals are available.
The passenger cabin is part of the Robotaxi equation
The R2’s purpose-built design gives DiDi more freedom to rethink the rear cabin around passengers, accessibility and fleet use.
The company highlights increased rear space, adjustable seats and a 17.3-inch ceiling-mounted display for entertainment functions.
An AI voice system can verify a passenger PIN, start the trip and control several cabin settings, DiDi says.
These features may improve convenience. Yet they also introduce questions about privacy, language support, accessibility and support during unexpected situations.
A convincing Robotaxi experience needs more than an empty driver’s seat. Passengers must understand what the vehicle is doing and how to seek help.
Why UAE testing could be more demanding
The UAE offers modern roads and strong institutional support for autonomous mobility. It also presents conditions that can challenge perception systems.
High temperatures, glare, dust, sand and fast-moving multi-lane traffic can all affect real-world performance and component durability.
Road markings, construction diversions and the behaviour of human drivers also vary from the environments used for Chinese testing.
Therefore, local trials should examine more than route completion. They should measure intervention rates, service interruptions and degraded-mode behaviour.
Cabin cooling, battery efficiency and sensor cleaning will also matter during long operating hours in the UAE summer.
DiDi would enter an active UAE Robotaxi market
The UAE is no longer discussing autonomous taxis only as a future concept. Dubai already has multiple operators and phased public services.

DiDi Autonomous Driving has also joined Abu Dhabi’s Smart and Autonomous Vehicle Industries (SAVI) cluster and established a strategic partnership with the Abu Dhabi Investment Office (ADIO) to support innovative autonomous-driving applications, AI talent development and ecosystem building.
Dubai’s Roads and Transport Authority reported more than four million autonomous kilometres and 7,613 passenger trips by 19 August 2026.
RTA also reported a 144-vehicle Robotaxi fleet, including 48 active vehicles, and a 97% customer satisfaction rate.
These figures describe Dubai’s wider Robotaxi programme. They do not represent DiDi operations or R2 performance.
Dubai aims to make 25% of transportation trips smart and driverless by 2030 across different transport modes.
DiDi has also joined Abu Dhabi’s Smart and Autonomous Vehicle Industries cluster, according to the supplied company announcement.
Its eventual role will depend on permits, local partnerships, service geography and evidence that the system suits UAE operating conditions.
The questions that matter before a UAE launch

- Where will the R2 operate, and what weather, road and speed limits will define its operational design domain?
- Will early UAE trials carry a safety operator, use remote assistance, or begin without anyone behind the wheel?
- What local safety data will regulators and the public receive before paid passenger journeys begin?
- How will emergency services, passengers and other road users communicate with an unoccupied vehicle?
- Where will trip, voice and cabin data be stored, and which privacy safeguards will apply in the UAE?
- How will the service handle extreme heat, dust accumulation, temporary diversions and inconsistent road markings?
CriticReviewer editorial perspective
The R2 trial moves DiDi from technical demonstration towards operating a passenger service under defined conditions.
That is meaningful progress, but it is not proof of global readiness. Geography can change the difficulty of autonomous driving dramatically.
For the UAE, DiDi’s strongest argument will not be its sensor count or computing figure. It will be transparent local performance.
The most useful next announcement would include test locations, safety supervision, operating limits and a clear path towards passenger access.
Until then, the R2 should be viewed as a promising candidate for UAE trials, rather than an imminent nationwide Robotaxi service.
Estimated reading time: 5 minutes
