Skip to content
Last Dictionary

subsumption architecture

nounAlso: subsumption, behavior-based robotics, behaviour-based robotics, reactive control

In plain wordsA way to build robot brains from stacked, reflex-like behaviors — a bottom layer that keeps clear of obstacles, others that roam or explore — where higher layers can take over from lower ones, with no central map or plan.

  1. A robot control architecture built from stacked layers of simple behaviors that run concurrently, each wiring sensing straight to action; a higher layer, such as one that explores, can take over from those beneath it, such as obstacle avoidance and aimless wandering, by suppressing their inputs or inhibiting their outputs. Layers are built and debugged bottom-up, so the robot stays useful if higher layers fail. The architecture deliberately avoids a central world model, symbolic representation and planner; Rodney Brooks summed up the stance as 'the world is its own best model'. It was a reaction against the sense–model–plan–act pipeline of classical AI robots such as ShakeyShakey the robot, prop. n.: A wheeled robot built at a California research institute between 1966 and 1972 that could look around, make a plan and carry it out: the first mobile robot to work out its own steps toward a goal., which were slow and brittle in real environments, and it founded behavior-based robotics, which produced fast, robust insect-like robots in the late 1980s and 1990s. Its limits on tasks needing memory, planning or abstraction led to hybrid architectures that place deliberative planning over reactive control, an arrangement echoed in robots that pair language-model planners with fast learned controllers.

    The cleaning robot's bump-and-turn behavior is subsumption in miniature.

Origin

Introduced by Rodney Brooks at MIT in 'A Robust Layered Control System for a Mobile Robot' (IEEE Journal of Robotics and Automation, 1986) and demonstrated on robots such as Allen, the six-legged walker Genghis and the can-collecting Herbert; Brooks developed the critique of representation in 'Elephants Don't Play Chess' (1990) and 'Intelligence without Representation' (1991), and co-founded iRobot, maker of the Roomba, in 1990.

Usage note

Mainly of historical and conceptual interest today, although layered reactive controllers remain common in practice and the debate it opened continues in embodied AIembodied AI, n.: AI that has a body, usually a robot and sometimes a simulated character, and learns or acts by sensing and moving in a world rather than only processing text or images..

Related

See also
  • embodied AIembodied AI, n.: AI that has a body, usually a robot and sometimes a simulated character, and learns or acts by sensing and moving in a world rather than only processing text or images.
  • Shakey the robotShakey the robot, prop. n.: A wheeled robot built at a California research institute between 1966 and 1972 that could look around, make a plan and carry it out: the first mobile robot to work out its own steps toward a goal.
  • mobile robotmobile robot, n.: A robot that travels around by itself — on wheels, tracks, legs, propellers or thrusters — such as a robot vacuum, warehouse cart, delivery robot or planetary rover.
  • good old-fashioned AIgood old-fashioned AI, n.: A nickname for the classic style of AI, dominant from the 1950s to the 1980s, that tried to make computers intelligent by writing knowledge and rules down by hand.
  • cognitive architecturecognitive architecture, n.: A blueprint for a complete thinking system — how memory, attention, learning and decision-making fit together — used both to build general AI agents and to model how the human mind works.
  • embodied cognitionembodied cognition, n.: The idea that thinking is not something the brain does on its own: the body, its senses and movements, and its surroundings play a real part in how we think and understand.
Contrast with
  • symbolic AIsymbolic AI, n.: An approach to AI that writes knowledge down as human-readable symbols and rules, such as 'every bird has wings', and reasons by manipulating them step by step instead of learning patterns from data.

Sources

  1. Brooks, R. A. (1986). A robust layered control system for a mobile robot. IEEE Journal of Robotics and Automation 2(1), 14–23.
  2. Brooks, R. A. (1991). Intelligence without representation. Artificial Intelligence 47, 139–159.