Search and Rescue Robot

In short: Building a search and rescue robot that could be operated intuitively via a VR headset and controllers.

Detailed: A Telepresence Control System allows for a more natural approach to controlling a complex robotic system. This is made possible by using a Virtual Reality headset and controllers. Virtual Reality technology is used to immerse a user into different worlds such as in games and 360° recorded videos. Telepresence control takes the technology one step further by using the parameters collected from the headset and hand controllers within a 3D space and mathematically modelling the position of each servo motor.

The search and rescue robot SAPHI engineers developed has two manipulator arms and a camera arm. The camera arm gives the pilot a first-person view of the position of the robot via a live low latency video feed. The VR headset roll, pitch and yaw sensors are used to determine the angles in which to move the servo motors for positioning the camera.  

The search and rescue robot’s two manipulator arms are controlled via VR hand controllers. The hand controllers are tracked relative to the position of the headset which produces data that can be used in a mathematical model of the manipulators. The model determines the angle in which 16 servo motors need to move to in real-time.

The manipulators have a set of pincers that are used to interact with the environment around the robot. The pincers position can be varied by the amount of pressure placed on triggers that are located on each hand controller. Haptic feedback pressure from the amount of torque being applied to the pincers is also fed back to the hand controllers in the form of vibration. The more pressure being placed on the item being pressed, the more the controller will vibrate.

The benefits of such a control system being used in a rescue situation include;

• Removing rescue personnel from dangerous situations
• Intuitively allow rescue personnel to interact with the environment surrounding the robot with minimal training