II. Mobile robotics

This part discusses mobile robots, a class of robots that are able to move through the environment: through the air, or through the water.

One of the most important functions of a mobile robot is to move to some place. That place might be specified in terms of some feature in the environment, for instance move to the light, or in terms of some geometric coordinate or map reference. In either case the robot will take some path to reach its destination and it faces challenges such as obstacles that might block its way or having an incomplete map, or no map at all.

One strategy is to have very simple sensing of the world and to react to what is sensed. For example Elsie the robotic tortoise (built in the 1940s reacted to her environment and could seek out a light source without having any explicit plan or knowledge of the position of the light. More recently robotic vacuum cleaners such as the iRobot Roomba follow a similar strategy.

An alternative to the reactive approach was embodied in the 1960s robot Shakey which was capable of 3D perception and created a map of its environment and then reasoned about the map to plan a path to its destination.

These two approaches exemplify opposite ends of the spectrum for mobile robot navigation. Reactive systems can be fast and simple since sensation is connected directly to action — there is no need for resources to hold and maintain a representation of the world nor any capability to reason about that representation. In nature such strategies are used by simple organisms such as insects. Systems that make maps and reason about them require more resources but are capable of performing more complex tasks. In nature such strategies are used by more complex creatures such as mammals.

The first commercial applications of mobile robots came in the 1980s when automated guided vehicles (AGVs) were developed for transporting material around factories and these have since become a mature technology. These free-ranging mobile wheeled vehicles typically use fixed infrastructure for guidance, for example, a painted line on the floor, a buried cable that emits a radio-frequency signal, or wall-mounted bar codes. More recently we have seen significant achievements in mobile robotics that can operate without navigational infrastructure. We have seen rovers on Mars, the DARPA series of grand challenges for autonomous cars and even small low-cost robotic vacuum cleaners. Field robotic systems such as trucks in mines, container transport vehicles in shipping ports, and self-driving tractors for broad-acre agriculture are now commercially available. Mobile robots are not just limited to operations on the ground and recent years have seen significant progress with unmanned aerial vehicles (UAVs), autonomous underwater vehicles (AUVs), and robotic boats (ASVs or autonomous surface vehicles).

  1. Mobile Robot Vehicles
    • Mobility
    • Car-like vehicles, moving to a point, line & pose
    • Flying robots
  2. Navigation
    • Reactive navigation, Braitenberg vehicles, Bug* automata
    • Distance transform, D*
    • Roadmap methods: Voronoi, PRM, RRT
  3. Localization
    • EKF-based dead reckoning
    • Map based
    • Creating a map
    • Localization & mapping
    • Monte-Carlo approach