1.MiCa’s deployment at Tallinn Airport marks a shift from testing autonomous mobility to using it in day-to-day airport operations. What made this transition possible, and what were the most significant challenges along the way?
The transition was possible because we approached it as a gradual operational and safety-validation process rather than a single technological demonstration.
We began testing at Tallinn Airport in 2023 with our first-generation autonomous shuttle, Iseauto. This allowed us to collect operational data, understand the airport’s procedures and validate our technology in real operating conditions.
MiCa entered the airport environment in 2024, and in 2025 it became part of Magnetic MRO’s commercial daily operations, transporting employees and equipment between the hangars. The shuttle operates seven days a week on an 11-kilometre round trip. In 2026, we added a second MiCa to the service, and today the two shuttles cover approximately 1,000 kilometres per week at the airport.
In August 2026, we received approval to operate without an onboard safety operator, and we demonstrated this capability publicly on 7 October. This made Auve Tech the first European company to operate a commercial autonomous passenger service at an airport without an onboard safety operator.
The most significant challenge was bringing together the technology, the operational procedures and the safety requirements of a live airport environment. An autonomous vehicle must function as part of the wider airport operation, alongside ground-support vehicles, aircraft movements and established access and safety procedures. This required close cooperation between Auve Tech, Magnetic MRO and Tallinn Airport, as well as continuous validation in the actual operating environment.
The removal of the onboard safety operator was therefore not one isolated step. It was the result of several years of testing, operational experience and progressively demonstrating that the service could be supervised safely from the remote supervision center.
2. Airports operate under strict safety requirements and involve multiple teams, vehicles and moving aircraft. Where can autonomous shuttles deliver the most operational value, and which airport processes are best suited to automation today?
The clearest value today is in repetitive, predictable transport tasks that take place on defined routes. These can include moving employees between terminals, hangars or other airport facilities, as well as transporting tools, spare parts and smaller equipment.
At Tallinn Airport, MiCa connects Magnetic MRO facilities across the airside area and supports the daily movement of employees and equipment. This is exactly the type of task where autonomy can create immediate value: the route is repeated, the operational area is clearly defined and transport is required consistently throughout the week.
Automating these journeys can reduce the amount of time specialised employees spend on internal transport and allows their expertise to remain focused on aircraft maintenance and other core activities. It can also make transport availability more consistent and reduce dependence on assigning a dedicated driver to every vehicle.
3. Removing the onboard safety operator is a significant step in autonomous transport. What technical validation and safety assessments were required before this could be authorized at Tallinn Airport?
The authorization was based on operational evidence gathered through testing and supervised operation on the actual airport route. The process included route-specific risk analysis, testing with a safety operator onboard, assessment of the vehicle’s performance and the definition of procedures for remote supervision and exceptional situations.
The route risk analysis considered airport-specific conditions such as aircraft right-of-way, intersections, unexpected obstacles, localization, lane width, emergency vehicles and occupied stopping positions.
The transition was completed progressively. Initially, the safety operator remained onboard while operational data was collected and the system was validated. Remote supervision was then introduced as an additional operational layer. Only after the relevant testing results and risk assessments had been reviewed was operation on the agreed route approved without an onboard safety operator, subject to remote supervision and continued review of operational results.
This phased process is essential. Removing the person from the vehicle does not mean removing human oversight from the operation. It changes where that oversight is located and requires the vehicle, the supervision system and the operational procedures to work together as one safety architecture.
4. Airports present a different operating environment from public roads, with restricted airside zones, ground-support vehicles, aircraft movements and changing conditions. How do you account for these risks when deploying autonomous vehicles in live operations?
We begin by defining the operational design domain and analysing the specific route rather than treating every airport as the same environment. This includes mapping the route, identifying interactions with aircraft and service vehicles, assessing intersections and stopping points, and defining how the shuttle must respond in each relevant situation.
MiCa uses onboard cameras and LiDAR sensors to perceive its surroundings and makes its driving decisions autonomously. The vehicle follows a predefined and validated route, while remote supervision provides an additional layer of operational oversight.
The deployment is also integrated with the airport’s operational rules. The vehicle must respect local right-of-way requirements, restricted areas, emergency procedures and coordination processes. Changes to the route or operating conditions cannot simply be treated as ordinary driving variations; they have to be assessed within the agreed operational framework.
The advantage of an airport is that, although it is safety-critical, it is also structured. Access is controlled, routes can be clearly defined and the movements around the vehicle are governed by established procedures. This makes it possible to build a precise, site-specific safety case.
5. What are the main limitations of remote supervision today? For example, how do connectivity failures, simultaneous vehicle alerts requiring human intervention affect the number of vehicles one operator can manage?
The main limitation is not simply how many vehicles can be displayed on one screen. It is how many vehicles one supervisor can oversee safely when several exceptional situations occur at the same time.
It is important to clarify what that assistance means. The supervisor does not take control of the shuttle or drive it remotely. There is no steering wheel or pedals at the remote supervision station. MiCa performs the driving task and path planning autonomously. If it encounters a situation where assistance is required, the supervisor can review the situation and give approval for a specific manoeuvre, which the vehicle then carries out autonomously.
Connectivity, alert frequency and the complexity of assistance requests all affect how many shuttles one operator can safely supervise. If communication is unavailable or the system cannot continue within its approved operating conditions, the vehicle must behave predictably and move into an appropriate safe state.
6. Tallinn Airport provides a real-world use case for autonomous mobility in a tightly regulated environment. Which aspects of this deployment can be replicated at other airports?
The individual route and approval process will always be site-specific, but the overall deployment method can be replicated.
This includes identifying a clear operational need, selecting a suitable repetitive route, mapping the environment, carrying out route-specific risk analysis and beginning operation with an onboard safety operator. Operational data can then be used to validate the system, develop local procedures and progressively introduce remote supervision. Once the required evidence has been established, the removal of the onboard safety operator can be considered with the relevant airport and authorities.
Airports, ports and industrial campuses are especially suitable because they often have controlled access, repeated internal journeys and clearly defined operating areas.
What can be replicated most directly is therefore not an identical route, but the progression from a clearly defined use case through testing, operational validation and phased removal of the onboard operator.
7. What is the biggest misconception airport operators still have about autonomous mobility, and what evidence should they examine before deciding whether to deploy it?
One misconception is that autonomous mobility should be assessed mainly as a technology demonstration. The more important question is whether it solves a recurring operational problem reliably and whether it can be integrated into existing airport procedures.
At Tallinn Airport, the key proof point is that MiCa supports Magnetic MRO’s regular transport of employees and equipment. The operation covers approximately 1,000 kilometres per week and has reported an average autonomy rate of 99.9%. By automating transport between the hangars, Magnetic MRO saves employees’ time that would otherwise be spent moving people and equipment around the airport. This allows specialised employees to focus more of their time on aircraft maintenance and other core tasks rather than internal transportation.
The misconception is that autonomous transport is expensive solution. The initial investment in an autonomous vehicle can be higher than in a conventional vehicle, but you have to look at the total operating cost rather than only the purchase price. The annual cost of the solution will be cheaper than conventional transport, due to needing less people to operate.
Airport operators should also examine whether the provider clearly distinguishes autonomous driving from remote driving. In our case, MiCa makes the driving decisions onboard. Remote supervision provides operational oversight and assistance when required; it is not a person continuously controlling the shuttle from a distance.
8. Looking ahead three to five years, what needs to happen across regulation, airport infrastructure and commercial adoption for autonomous vehicles to become a standard part of airport operations?
First, regulation needs to catch up with the technology. In many markets, there is still no clear legislation specifically covering autonomous vehicle operations. More established regulatory frameworks would make it easier to move from individual approvals and deployments toward wider adoption.
Second, airports need confidence in the security of autonomous vehicles. This includes both the general security and safety of the vehicle and cybersecurity. Autonomous vehicles are connected systems, so demonstrating that the vehicle, its communications and the wider system are secure is an important part of building confidence in large-scale deployment.
Third, commercial scalability will depend heavily on remote supervision. We have seen that the point where autonomous transport starts to become a competitive alternative to conventional transport is when one supervisor can oversee multiple vehicles. Depending on the operating environment, this can already be achieved at ratios of 1:2 or 1:3. Our longer-term objective is 1:10, where one supervisor can oversee up to ten autonomous vehicles. Increasing the number of vehicles one supervisor can oversee changes the economics of the service significantly, because the fleet can grow without the number of operators having to grow at the same rate.
Auve Techs shuttle MiCa demonstrates that autonomous mobility can already move beyond a pilot project and become part of everyday airport logistics. The next step is to turn the experience gained from individual deployments into a scalable and repeatable operating model for other airports.
Author Bio:
Johannes Mossov is a member of the Management Board at Auve Tech, supporting the company’s strategic growth and business development. With over 10 years of experience in autonomy, he has extensive expertise in custom vehicle manufacturing and broad know-how from various manufacturing teams, helping drive the commercialization of Auve Tech’s autonomous mobility solutions.
Company Info:
Auve Tech is an Estonian developer and manufacturer of autonomous vehicles and software for airports, industrial sites, campuses and urban environments. Founded in Tallinn, its shuttles operate internationally and provide safe, sustainable and cost-effective mobility through autonomous driving and remote supervision.
