Hanoi is not yet launching a pilotless air-taxi service. It is studying a controlled regulatory sandbox for the EH216-S at Hoa Lac High-Tech Park, and the central test is not whether the aircraft can take off. It is whether Vietnam can build an accountable system around an aircraft that carries people without a pilot on board.
HungViet Technology and Investment Trading has signed a memorandum of understanding with Hanoi’s Department of Science and Technology to research and propose the sandbox. EHang is to provide the pilotless aircraft, operating technology and training support. HungViet is expected to provide local technical, financial and human resources, while the Hanoi authority guides the application and approval process. Controlled flights may begin only after formal approval.
That distinction is important. An agreement to pursue a trial is not permission to carry fare-paying passengers over Hanoi. The trial remains limited to Hoa Lac High-Tech Park, and city officials say documentation, procedures and formal approval must be completed before the aircraft can fly under the proposed sandbox.
The EH216-S is a two-seat, electric vertical-take-off-and-landing aircraft designed to fly autonomously. EHang says it has received type, production and standard airworthiness certificates from the Civil Aviation Administration of China. Those approvals matter, but they do not automatically transfer to Vietnam, India or any other jurisdiction. Each regulator must decide whether the aircraft, operating concept, maintenance process, communications links and safety case meet local requirements.
That is why a sandbox is a governance experiment. An ordinary flight safety question asks whether an aircraft is airworthy. A pilotless passenger flight asks a longer list of questions. Who observes the aircraft. Who decides whether weather is acceptable. Who monitors the battery and propulsion system. Who intervenes if navigation becomes unreliable. Who owns the operational decision to continue, divert or land. Who is liable if the autonomy system makes the wrong choice.
A pilot has traditionally carried immediate responsibility for the aircraft’s safe operation. In a pilotless eVTOL, that responsibility becomes a chain.
The manufacturer may be responsible for design, software updates and known technical defects. The operator may be responsible for maintenance, dispatch, passenger safety and operational control. A remote supervisor may be responsible for monitoring several aircraft. A vertiport operator may be responsible for landing-site safety. The regulator must decide what must be reported, who can suspend operations and what level of autonomy is acceptable.
If something fails, the public will not accept an answer that says responsibility was shared too widely to locate. The aircraft therefore needs more than a route. It needs a safety envelope.
That envelope includes segregated or carefully managed low-altitude airspace, geofencing, weather limits, emergency landing areas, battery reserves, communications redundancy, cyber protection, maintenance records, secure software configuration and procedures for passengers who may be unable to assist in an emergency.
A pilotless aircraft cannot simply descend onto any open field when something goes wrong. It needs pre-assessed sites that account for people, power lines, roads, buildings, weather, access for emergency services and the ability to secure the aircraft after landing.
Air-traffic management for conventional aircraft is designed around pilots, controllers, radio procedures, flight plans and established separation rules. A low-altitude autonomous network will need to combine aircraft position, battery condition, weather, temporary restrictions, obstacle data, emergency alerts and interactions with helicopters, drones and conventional aviation.
The system must know not only where an eVTOL is. It must know whether it is still capable of finishing the flight safely.
Cybersecurity is central because autonomy moves some risk from the cockpit into software, networks and data. An aircraft’s flight-control logic, navigation data, communications link, remote monitoring station and maintenance laptop can all become targets for error or malicious interference. A credible sandbox needs penetration testing, secure update procedures, authentication controls, incident reporting and a clear rule for what happens when communications degrade.
The easiest answer is often to say the aircraft will land automatically. The harder question is where, under what conditions and with what assurance that an automated emergency landing does not create a new hazard on the ground.
Insurance will force the same questions into financial terms. Insurers will want evidence about failure rates, software assurance, battery safety, maintenance compliance, third-party risk, passenger liability, cyber exposure and emergency response. A manufacturer’s certification is important, but an insurer will also ask how the particular operator intends to use the aircraft in the particular city.
Who regulates low-altitude autonomous passenger operations. How do eVTOL routes interact with airport airspace, helicopter routes, military zones and commercial drones. What certification standard applies to aircraft and batteries. Who operates vertiports. What emergency landing locations are available in dense Indian cities. How will police, fire services, hospitals and air-traffic authorities coordinate after an incident.
Hanoi’s sandbox is valuable precisely because it does not pretend that a pilotless air taxi is merely a smaller helicopter. It is an attempt to test an aircraft, a traffic-management system, an operations centre, maintenance processes, public acceptance and legal responsibility together.
The real question is not whether the EH216-S can fly without a pilot. It is whether a city can know when it should not.