Skip to main content

Locomotion and Mobility

This chapter explores the fundamental principles of locomotion in robotics, covering various approaches to enabling robots to move through their environment. Locomotion is a critical capability for Physical AI systems that must navigate and interact with the physical world.

Introduction to Locomotion​

Locomotion is the ability to move from one location to another. In robotics, locomotion systems must balance multiple competing requirements: stability, efficiency, speed, adaptability, and robustness. The choice of locomotion strategy significantly impacts the robot's design, control complexity, and application domains.

Locomotion Classification​

Robotic locomotion can be classified by:

  • Contact with ground: Continuous (wheeled, tracked) vs. intermittent (legged)
  • Stability: Static (stable at all times) vs. dynamic (stable in motion)
  • Terrain adaptability: Specialized vs. general-purpose
  • Energy efficiency: Optimal vs. versatile

Wheeled Locomotion​

Wheeled locomotion is the most common form of ground-based mobility due to its efficiency and simplicity.

Wheel Configurations​

Standard Configurations​

  • Differential Drive: Two independently controlled wheels with a caster for balance

    • Simple control (linear and angular velocity)
    • Non-holonomic (cannot move sideways)
    • Common in mobile robots and wheelchairs

  • Ackermann Steering: Four wheels with steerable front wheels

    • Used in cars and similar vehicles
    • Efficient for high-speed operation
    • Complex kinematics and dynamics

  • Mecanum Wheels: Special wheels with rollers at 45° angles

    • Holonomic motion (can move in any direction)
    • Complex mechanics and control
    • Useful for precise positioning

Advanced Configurations​

  • Omni-wheels: Wheels that can move in any direction
  • Spherical wheels: Theoretical ultimate omni-directional motion
  • Tracked systems: Continuous tracks like tanks

Wheeled Robot Kinematics​

Differential Drive Kinematics​

For a differential drive robot with wheel separation L:

v = (vᵣ + vₗ) / 2
ω = (vᵣ - vₗ) / L

Where v is linear velocity, ω is angular velocity, and vᵣ, vₗ are right and left wheel velocities.

Legged Locomotion​

Legged locomotion provides exceptional terrain adaptability but at the cost of increased complexity.

Legged Robot Classification​

By Number of Legs​

  • Biped: Two legs (human-like)

    • High terrain adaptability
    • Complex balance control
    • Energy-intensive

  • Quadruped: Four legs (dog-like)

    • Natural stability
    • Good balance and speed
    • Complex coordination

  • Hexapod: Six legs (insect-like)

    • High stability (static or dynamic)
    • Complex kinematics
    • Excellent terrain adaptability

  • Multi-legged: More than six legs

    • Very high stability
    • Extremely complex control
    • Often used in research

Gait Analysis​

Static vs. Dynamic Stability​

  • Static stability: Center of mass remains within support polygon at all times

    • Safe but slow
    • High energy consumption
    • Simple control

  • Dynamic stability: Stability maintained through motion

    • More efficient
    • Faster locomotion
    • Complex control

Common Gaits​

  • Tripod Gait (hexapod): Three legs move while three support

    • Fast and stable
    • Good for level terrain

  • Ripple Gait (hexapod): Alternating diagonal pairs

    • More stable than tripod
    • Slower than tripod

  • Wave Gait (hexapod): Sequential leg movement

    • Most stable
    • Slowest gait

  • Walk Gait (quadruped): Three legs support while one moves

    • Stable and energy-efficient
    • Slow but safe

  • Trot Gait (quadruped): Diagonal pairs move together

    • Faster than walk
    • Moderate stability

  • Gallop Gait (quadruped): Both hind legs move first

    • Fastest gait
    • Dynamic stability required

Balance Control​

Zero Moment Point (ZMP)​

ZMP is a critical concept in legged locomotion stability:

ZMP_x = (Σ(F_z * x - M_y)) / ΣF_z
ZMP_y = (Σ(F_z * y + M_x)) / ΣF_z

Where F_z are vertical forces, M_x, M_y are moments, and (x,y) are force locations.

Center of Mass Control​

Maintaining the center of mass within the support polygon is essential for stability:

  • Capture Point: Location where the robot can come to rest
  • Linear Inverted Pendulum Model: Simplified model for balance control
  • Model Predictive Control: Advanced approach for dynamic balance

Aerial Locomotion​

Aerial robots (drones) provide unique mobility capabilities but require different control approaches.

Multirotor Configuration​

  • Quadrotor: Four rotors, simplest configuration
  • Hexarotor: Six rotors, redundant capability
  • Octocopter: Eight rotors, heavy payload capability

Control in Aerial Systems​

Aerial robots control four degrees of freedom:

  • Vertical thrust
  • Roll, pitch, yaw torques
  • Complex aerodynamics and control

Swimming and Underwater Locomotion​

Underwater robots face unique challenges:

  • Buoyancy control
  • Pressure resistance
  • Propulsion in fluid medium
  • Communication limitations

Hybrid Locomotion​

Some robots combine multiple locomotion modes:

  • Wheeled-legs: Wheels for efficiency, legs for obstacles
  • Ground-air: Ground locomotion with flight capability
  • Multi-modal: Several locomotion modes for different terrains

Terrain Adaptation​

Rough Terrain Navigation​

  • Path planning: Finding traversable routes
  • Gait adaptation: Changing locomotion pattern
  • Compliance: Adapting to surface irregularities

Dynamic Terrain Interaction​

  • Slip modeling: Understanding wheel-ground interaction
  • Adaptive control: Adjusting for changing conditions
  • Learning: Improving performance over time

Mobility Control Algorithms​

Path Following​

  • Pure Pursuit: Follows a sequence of way points
  • Stanley Controller: Corrects for cross-track error
  • Model Predictive Control: Optimizes over prediction horizon

Obstacle Avoidance​

  • Local planning: Real-time obstacle avoidance
  • Reactive methods: Immediate response to obstacles
  • Potential fields: Virtual forces for navigation

Formation Control​

For multiple robots:

  • Leader-follower: One robot leads others follow
  • Virtual structure: Robots maintain geometric pattern
  • Behavior-based: Emergent formation from local rules

Performance Metrics​

Mobility Metrics​

  • Speed: Maximum and average velocity
  • Efficiency: Energy per distance traveled
  • Payload capacity: Weight carried vs. robot weight
  • Terrain capability: Types of terrain traversable
  • Autonomy: Operating time without recharging

Stability Metrics​

  • Stability margin: Distance from stability boundary
  • Recovery time: Time to recover from disturbances
  • Robustness: Performance under uncertainties

Challenges and Considerations​

Energy Efficiency​

  • Actuator efficiency: Converting energy to motion
  • Transmission losses: Energy lost in mechanical systems
  • Control optimization: Minimizing unnecessary motion

Environmental Factors​

  • Terrain variability: Adapting to changing ground conditions
  • Weather effects: Wind, rain, temperature impacts
  • Obstacle density: Navigating cluttered environments

Control Complexity​

  • Real-time constraints: Control updates within time limits
  • Sensing requirements: Accurate state estimation
  • Computational demands: Processing power for control

Chapter Summary​

This chapter covered the fundamental principles of locomotion in robotics, from wheeled systems to legged robots and aerial vehicles. Understanding locomotion is crucial for designing robots that can effectively navigate and interact with their environment in Physical AI applications.

Exercises​

  1. Analysis Exercise: Compare the advantages and disadvantages of legged vs. wheeled locomotion for traversing rough terrain. Consider factors such as energy efficiency, speed, stability, and complexity.

  2. Design Exercise: Design a gait pattern for a quadruped robot that balances stability and speed. Specify the foot placement sequence and timing constraints.

  3. Implementation Exercise: Implement a simple balance controller for a biped robot using center of mass control.

Review Questions​

  1. What is the difference between static and dynamic stability in legged locomotion?
  2. Explain the Zero Moment Point (ZMP) concept and its importance in balance control.
  3. What are the advantages of differential drive vs. Ackermann steering?
  4. Describe the different gaits for quadruped robots and their characteristics.
  5. What are the main challenges in aerial robot locomotion?

References and Further Reading​

  • [1] Raibert, M. (1986). Legged Robots That Balance.
  • [2] Lynch, K. M., & Park, F. C. (2017). Modern Robotics: Mechanics, Planning, and Control.
  • [3] Hirose, S., & Takeuchi, H. (2001). Biologically Inspired Robots: Snake-like Locomotors and Manipulators.