The Mobile Advanced Robotics Defence Initiative (MARDI) was a United Kingdom Ministry of Defence (MoD) programme initiated in 1989 at the Royal Armament Research and Development Establishment (RARDE) in Chertsey, Surrey.

In the early eighties, the Military Vehicle Experimental Establishment started work on robotic concepts that would be realised later that decade in MARDI, building on work completed for explosive ordnance disposal applications in Northern Ireland and the Falkland Islands

MARDI involved partnerships with industry (including Alvis, British Aerospace, CAP Scientific, GKN Defence, and Lucas Applied Technology) and academic institutions to advance autonomous and teleoperated robotic ground vehicles for military applications.

Building upon earlier efforts such as the Vehicle Electronics Research Defence Initiative (VERDI), MARDI aimed to bring together and develop battlefield robotics through improved autonomy, sensor integration, and vehicle platforms suitable for complex environments.

The programme operated through the early 1990s, with key demonstrations in 1991, and influenced subsequent UK defence robotics research by focusing on reducing human intervention in hazardous operations.

MARDI – Key Technical Components #

MARDI centred on practical demonstrations of robotic technologies for combat scenarios, including reconnaissance, surveillance, and logistics support, exploiting the building blocks that the Military Vehicles and Engineering Establishment (MVEE) and others had been working on since the early eighties.

Autonomy Levels and Control Architectures #

MARDI demonstrated “bounded autonomy” and teleoperation, where robots executed specific tasks with minimal adaptability, progressing towards supervised autonomy for more generalised operations. This involved hierarchical control systems that allowed transition from manual teleoperation (human-in-the-loop for real-time guidance) to partial autonomy with onboard intelligence for path execution and decision-making under uncertainty.

Development of adaptive control strategies incorporating kinematics, dynamics, and mobility models for structured and unstructured terrain. This ensured stability and manoeuvrability, reducing operator workload in dynamic environments. For instance, algorithms for path planning and obstacle avoidance were tested, enabling robotic vehicles to navigate with reduced human input.

Sensor Integration and Environmental Perception #

Robust and redundant sensor suites were a core achievement of MARDI, including video cameras and processing units for obstacle detection, path planning, and real-time environmental mapping.

Multi-sensor fusion techniques were also implemented to enhance reliability, countering issues like deception or jamming through complementarity, timeliness, and cost-effective redundancy.

Vision systems developed and tested include pan-tilt cameras with photogrammetry, force/torque sensors, tactile grippers, and acoustic sensors for data integration. Loose and tight coupling methods (external data structures or shared memory) facilitated sensor data management, supporting self-organisation and adaptation (e.g., in case of sensor failure).

Vehicle Platforms #

MARDI utilised a modified tactical platform, the Alvis Streaker primarily

The MARDI Streaker had a full surveillance system including laser rangefinder, thermal camera and two daylight cameras (one for driving and the other on its sensor mast)

It also had a smoke generator, one of the applications envisaged was smokescreen generation.

The 4 tonne Bedford truck in the background of the image above was the control cabin, although it could be operated up to 6km away.

MARDI also used a small UGV called HARP, also part of VERDI.

Field tests validated improved mobility, with vehicles capable of teleoperated or semi-autonomous navigation in hazardous settings.

Given the available technology at the time, MARDI was an influential development programme, the remote control system was eventually utilised in the Terrier combat engineering vehicle, for example.

If you found value in this article, help me keep Think Defence going.

---

---