The first and most critical step in building a robotic system is selecting the platform. This is usually dictated by application specific mobility requirements, such as terrain, distance or payload as well as the capabilities and mission behaviors. Once the platform has been selected, it must be converted to drive-by-wire control so that all the functions necessary to safely operate the vehicle can be controlled and monitored electronically, preferably through a unified interface using a documented control standard. It most cases, this is more than just throttle, brake and steering, but also includes controlling engine status, fuel levels, lights, signals, horn, wheel speeds and more.
While safety should be considered in all aspects of robotic system design and operation, there are some additional provisions that need to be incorporated. Safety systems protect those operating near the vehicle as well as the vehicle itself. All robotic systems should incorporate an emergency stop system. This should be tied to the lowest level of vehicle control, making use of hardware only solutions whenever possible. The emergency stop should be controlled by on-vehicle stop buttons as well as an independent, wireless emergency stop system. Robotic systems should also have an emergency manual override system in place when a manned safety driver is present.
The first challenge in powering a robotic system is to determine the source of the power, which in most cases is preferable to have the tied directly to the vehicle, for reliability. Once the source is identified, it is necessary to determine the types and quantities of power required, which is largely based on the complexity of your system and payloads being added. Due to the unmanned nature of robotic systems, it is necessary to have an intelligent power distribution system that can be controlled and monitored electronically to allow for control of subsystem power from the operator control unit, the autonomous system, or for health monitoring and recovery.
The key to making an unmanned vehicle autonomous is the integration of complex autonomous navigation software. The level of complexity depends on the capabilities and behaviors required for your application, which determine the sensing, perception and computing power needed. Autonomous navigation could range form simple point-to-point navigation on a flat, open road up to dynamic obstacle avoidance and fully autonomous operation in rugged off-road environments.
Whether your robotic system uses teleoperated or autonomous control, there is a need for an operator to interface with the robot. Teleoperated control requires the operator to have direct input to the system through hands on controls such as a joystick, video game style controller, or simulated vehicle interface. Autonomous control typically consists of pre-mission planning, which may consist of sending a target waypoint, planning a route or generating a priori knowledge of the operating environment, and supervisory control during operation, such as adjusting goal points, monitoring status and providing high level input and control.