TIER IV, Taisei Corporation and Taisei Rotec have been selected for a Japanese government demonstration examining how autonomous vehicles could operate within dedicated freight transport infrastructure. The project will test three experimental vehicles in controlled tunnel and road environments, with the results intended to inform the development of Japan’s proposed Autoflow Road logistics system.
The Ministry of Land, Infrastructure, Transport and Tourism has commissioned the companies to investigate the integration of vehicle control, localisation and roadside equipment. An eight-day programme is scheduled between late October and early November 2026 at facilities operated by the ministry’s National Institute for Land and Infrastructure Management.
Testing will use a full-scale tunnel facility and sections of an associated test track. Planned manoeuvres include vehicle following, stopping, passing oncoming vehicles, merging, diverging, avoiding obstacles and approaching designated stopping positions. Engineers will also examine the space required to perform those movements within different lane widths.
The three experimental vehicles are based on passenger cars. They will be used to investigate autonomous driving and coordination rather than carry commercial freight, and the companies have made clear that the test fleet does not represent the dedicated logistics equipment that could eventually operate on an Autoflow Road.
Japan’s proposed system would move cargo through specially constructed transport corridors using unmanned vehicles. Separating those vehicles from conventional traffic creates opportunities to control operating conditions more closely, while introducing substantial engineering requirements for the infrastructure through which they move.
Ordinary roads contain unpredictable interactions involving other vehicles, pedestrians, roadworks and changes in traffic control. A dedicated logistics corridor could reduce some of that variation, but its equipment must still respond reliably to other autonomous vehicles, stationary obstacles and changing conditions within the route.
The demonstration is examining how responsibility should be divided between the vehicle and the surrounding infrastructure. A conventional autonomous vehicle may carry numerous sensors to determine its position and interpret the environment. In a purpose-built corridor, some functions could potentially be supported by installed roadside equipment.
Reducing the complexity of individual vehicles could affect the cost of a future freight system. However, transferring functions to infrastructure changes where maintenance and reliability responsibilities sit. Equipment installed along a route must remain available and provide sufficiently accurate information whenever vehicles depend on it.
Vehicle localisation is central to the programme. An automated system must determine its position relative to the intended driving path before its controllers can calculate steering and speed adjustments. Errors in that estimate can affect the available clearance from neighbouring vehicles or fixed structures.
Satellite positioning can support localisation in open environments, but its availability is restricted in tunnels and other enclosed spaces. Vehicles must therefore rely more heavily on other measurements, which can include onboard sensors, inertial information and features within the surrounding infrastructure.
Taisei Corporation has developed T-Localizer, a technology intended to support positioning through infrastructure-based references. Its performance will be assessed during the forthcoming trials, including the potential to simplify the sensors installed on individual vehicles.
Taisei Rotec will undertake reflector installation planning, construction and measurement work. The company has also developed tunnel surface treatments intended to improve how lidar sensors recognise features in environments where satellite positioning is unavailable.
Lidar determines distances by measuring the return of emitted light from surrounding surfaces. The reliability of those measurements depends on the sensor and the characteristics of the surfaces being observed. Suitable reference features may improve localisation, although their arrangement, condition and visibility must be considered within the complete system.
Reflectors and surface treatments also create maintenance requirements. Contamination, damage or changes to the physical environment could affect the references available to a vehicle. A practical installation needs arrangements for monitoring such conditions and responding when the information required for safe operation becomes unreliable.
Lane width is another focus of the project. Engineers will assess standard, intermediate and minimum driving spaces to establish the clearances required for planned manoeuvres. The width of a corridor influences construction requirements, but it also determines how much room vehicles have for steering corrections and interactions with other equipment.
Passing and merging operations create particular constraints because two vehicles may need access to overlapping parts of the route. Their control systems must determine appropriate movement sequences and maintain the separation required by the operating conditions.
The demonstration will test following and stopping behaviour involving multiple vehicles. These operations depend on the system’s ability to detect or receive information about the speed and position of nearby equipment and respond when movement changes.
Obstacle avoidance will be examined using objects intended to represent material obstructing the route. A vehicle may need to stop or move around an obstruction, but the available response depends on the corridor width, other vehicles and the position of the obstacle.
Stopping accuracy is important for a future logistics operation because vehicles may need to align with loading equipment or defined transfer positions. Errors that would be tolerable during ordinary movement could prevent reliable transfer of cargo at a fixed station.
TIER IV will supply the experimental vehicles and Pilot.Auto, its autonomous driving system based on the open-source Autoware platform. The company will support driving operations and collect information needed to assess vehicle behaviour during the different manoeuvres.
Taisei Corporation is responsible for overall planning, localisation technology assessment, infrastructure simplification and evaluation of the results. Taisei Rotec’s contribution concentrates on the physical infrastructure and field measurements needed to support the experiments.
The three organisations previously participated in a government demonstration during the 2025 financial year. That work examined T-Localizer and Taisei’s T-iPower Road wireless power transfer technology, identifying potential applications as well as operating issues requiring further investigation.
Wireless transfer could eventually provide electricity to suitable vehicles through equipment incorporated into a transport route. Its practical performance would depend on alignment, transfer efficiency, operating speed and compatibility between the installed infrastructure and the receiving equipment.
The October announcement does not establish that wireless charging will form part of a commercial network or that the forthcoming tests will demonstrate a fully operational freight transport system. The current programme concentrates on autonomous movement, vehicle coordination and the balance between onboard and roadside equipment.
The participants have not announced a commercial route, planned carrying capacity, construction budget or launch date for an operational Autoflow Road. Those decisions would require additional evidence about the vehicles, loading systems, infrastructure costs and reliability of continuous operation.
The eight-day demonstration is intended to establish which vehicle and infrastructure arrangements can perform the required manoeuvres within controlled test conditions. Its results will inform the technical design of future logistics corridors, but further testing with purpose-built cargo equipment and sustained operating cycles will be necessary before commercial deployment can be assessed.
