The Mindsheet Team

Team Name: Testudo

The concept is named Testudo after the Roman legion protective battle formation that looked like a tortoise.

Roman army in Testudo Formation

Team Leader: Raglan Tribe

Figure 1: Team Leader - Raglan Tribe

Figure 1: Team Leader - Raglan Tribe

Team Location: Havant, Hampshire

Team members

picture of the Mindsheet team outside their headquarters

Chris Burgess, Raglan Tribe, Ken Maclauchlan, Catherine Dollery, Paul Freeman, Susan Tribe, Andy Hubbard, Don Payne, Harbir Singh, Steve Brunswick, Colin Frey and John Tribe. Not shown: Phil Cooper.

Sponsors

Mindsheet Ltd is sponsoring the Testudo team. Mindsheet specialises in helping companies develop breakthrough products and services in highly complex environments where understanding how to deliver value to customers requires deep insight into their operations and behaviour. Mindsheet has the resources and capabilities to rapidly take breakthrough ideas through the full development lifecycle from concept development to system integration and testing.

Outline System Concept

Figure 2: Testudo System Elements

Figure 2: Testudo System Elements

The Testudo Concept relies on the cooperation between multiple autonomous robot platforms that can detect, identify, locate and report a range of threats in a hostile urban environment as shown in Figure 2. The main system elements are a:

  • Ground-based autonomous vehicle which is capable of following a preset GPS trajectory with local obstacle avoidance by the use of ultrasonic sensors. On the back of the vehicle are video cameras that are pointing at the windows on the sides of the houses. The video is transmitted to a relay station on the helikite and then down to the ground station.
  • Aerial Surveillance Platform and Relay station a helikite is used to provide a stable platform for mounting thermal and visible cameras which are trained on the designated threat zone in the town. Again, the video signals are transmitted down to the ground station for digital signal processing and threat monitoring. Additionally, the helikite relays transmissions from the autonomous vehicles down to the ground station.
  • Ground Station with the automatic threat detection each video feed is processed by a dedicated computer and the threat identity and location is sent to a PC laptop upon threat recognition. The core part of the innovation is the recognition and data fusion algorithms that can classify and locate threats in complex 3-D environments from 2-D imaging sensors. At the heart of the threat recognition are hierarchical temporal memory systems that can learn threats in moving imagery based on the identification of human poses as shown in the Figure 3.
Figure 3: Threat identification by pose recognition of video images

Figure 3: Threat identification by pose recognition of video images