Development documentation

• Product specification
  • A basic bot
  • A better bot
  • A good bot
  • A great bot
  • A dream bot
• Dr.QP CAD designs
  • Web viewer
  • Downloads
• Kinematics model
  • 1. Getting Started With Robot Inverse Kinematics (IK)
    • What is Inverse Kinematics?
    • Why is Inverse Kinematics Important?
    • How to Use Inverse Kinematics
    • The robot model
    • Forward kinematics
    • Exercise 1. Forward kinematics. Find angles at which the leg is on the ground
    • Solving inverse kinematics using trigonometry
      • Coxa Inverse Kinematics
    • Femur and Tibia Inverse Kinematics in X’Z Plane
      • Putting it all together
      • The classic IK test - drawing a circle
  • 2. Matrix based transforms
    • Affine transforms
  • 3. Generating gaits
    • Parametric gait generator
    • Visualizing the gaits
      • Animating the gaits
      • Wave gait
      • Ripple gait
      • Tripod gait
    • Directional Gait Decorator
      • TL;DR
      • Why This Works
      • Putting it all together
  • 4. Migrating the Existing Kinematics Model to MoveIt 2
    • What existed before MoveIt 2?
    • The old mental model in one picture
    • Step 1. Preserve the robot’s structure in URDF/Xacro
      • 1.1 The body file instantiates six legs
      • 1.2 The leg macro defines the same three actuated joints
      • 1.3 Why there are many fixed joints inside one leg
    • Step 2. Add the semantics MoveIt needs through SRDF
      • 2.1 Planning groups
      • 2.2 End effectors
      • 2.3 Named states
      • 2.4 Disabled collision pairs
    • Step 3. Replace the handwritten IK entry point with a MoveIt solver plugin
    • Step 4. Carry over limits and execution constraints
      • 4.1 Joint limits
      • 4.2 Controller mapping
    • Step 5. Assemble everything in the launch files
    • Step 6. Support both visualization-only and simulation flows
      • 6.1 RViz-only demo
      • 6.2 Gazebo demo
    • Step 7. Verify the migration with automated tests
    • Old world vs new world
    • What did not change
    • A practical migration checklist
    • Summary
  • Experiments
    • SAxense feasibility for DualSense audio and haptics
      • Executive summary
      • What SAxense actually does
      • Why SAxense is the right tool for Bluetooth haptics on Linux
      • Integration paths
        • Option A — ffmpeg with pre-recorded clips (recommended for Dr.QP)
        • Option B — PipeWire persistent sink
      • Existing Dr.QP infrastructure fit
      • Estimated latency budget
      • Known issues with PipeWire on Ubuntu 24.04 / Raspberry Pi 5
      • Open issue in SAxense issue tracker
      • Acceptance criteria status
      • Recommended integration approach
        • Architecture
        • Design constraints
        • Build additions needed
        • Process lifecycle
      • Failure modes and mitigations
      • What must happen before code ships
      • Recommendation
      • Alternatives considered and rejected
    • 3.1 Experiment: Smoothing Splines
      • Deep dive into smoothing splines
      • BSpline
      • BSpline in application to gaits trajectory approximation
      • Bezier curve using B-spline
      • Pure python implementation of Bézier curve
        • Bezier Interpolation
      • An abandoned attempt to use smoothing splines for gait generation
        • Summary and steps forward (pun intended)
        • Gait trajectory function
      • Using splines for trajectory interpolation
        • Defining gaits as b-spline control points
        • Attempt to interpolate the trajectory
    • 3.2 Experiment: Transitions
• Notes
  • Raspberry Pi storage
    • Tested on RPi 4
      • Built in SSD (Ultra XC10)
      • Samsung tiny USB3 Drive (FAT)
      • Samsung tiny USB3 Drive (ext4)
      • SanDisk via proper USB3 cable
      • Kingston DataTraveler MAX
    • Tested on linux laptop
      • Kingston DataTraveler MAX via Dock → Hub
      • Kingston DataTraveler MAX directly via USB3 adapter directly
  • Uninstalling Ubuntu Desktop
  • vcpkg evaluation