In this lesson, we’ll explore the architecture of autonomous vehicles, which includes the critical components that allow them to perceive, plan, and operate safely and efficiently.

  1. Introduction
  • The previous chapter discussed how autonomous vehicles perceive their environment through techniques like localization, mapping, and object recognition.
  • This lesson focuses on how autonomous vehicles combine this environmental knowledge with other data to safely navigate to their destination.
  • Autonomous vehicle software can be divided into two main perspectives: perception and planning/operation.
  • The architecture combines these functions to achieve the desired level of autonomous driving.
  1. Perception
  • Perception aims to answer the questions “where am I?” and “what’s happening around me?”
  • Key functions of perception include localization, mapping, and object recognition.
  • These functions were explained in detail in the previous chapter and serve as the foundation for decision-making.
  1. Planning
  • Planning addresses the question “how do I get to my destination?”
  • Planning activities follow a three-layer hierarchy: route planning, behavior planning, and movement planning.
  • Route planning defines the overall path to the destination.
  • Behavior planning determines how the vehicle should behave in response to its surroundings and the desired route.
  • Movement planning translates these behaviors into specific actions for the vehicle to execute.
  1. Operating the Vehicle
  • The vehicle control aspect of the architecture is responsible for executing the decisions made in the planning stages.
  • Its primary role is to ensure the safe movement of the vehicle.
  • Vehicle control includes translating calculated paths into control commands for the vehicle’s actuators.
  • It focuses on maintaining vehicle stability and mitigating the impact of unexpected events.
  • Safety is a paramount concern, and control modules often function as redundant safety systems.
  • These modules can override higher-level decisions to prevent accidents or reduce their severity.
  • Vehicle control also handles steering the vehicle in both lateral and longitudinal directions.
  • Tasks include lane-keeping, speed control, maintaining safe distances from other vehicles, and lane changes.


  • Autonomous vehicle architecture is a complex system that integrates perception, planning, and vehicle control.
  • Perception gathers data about the environment, while planning decides on the route and behavior.
  • Vehicle control executes these decisions, ensuring safe and efficient movement.
  • Safety and redundancy are critical aspects of autonomous vehicle control, preventing accidents and maintaining operational integrity.
  • Understanding this architecture is essential for anyone involved in the development and research of autonomous vehicles.
Go back