Examples

Decision tree — which example should I read?#

Are you new to the framework?
├── YES → start with the atomic examples (one concept each).
│         See [Atomic examples](atomic.md) — 5 single-file demos
│         covering Steps 1, 2, 3, 4, and 6/7 of the
│         implementation guide.
│
└── NO → what does your application look like?

       Continuous joint-angle control / antagonist muscle pairs?
       └── [Elbow flexion](elbow_flexion.md) — PID control of
           biceps + triceps with auto-tuner and PyQt5 GUI.

       Haptic / tactile feedback (non-functional EMS)?
       └── [Haptic navigation](haptic_navigation.md) — directional
           cues via `HapticLibrary`.

       Need a GUI scaffold without a real strategy?
       └── [Generic GUI](generic_gui.md) — reference PyQt5 template
           wired to the framework's `SharedRingBuffer`.

       Gait-phase-adaptive walking FES?
       └── [Walking FES](walking_fes.md) — the
           reference application; complete with ML step detector,
           per-leg sliders, real-time biomech plots.

Catalog#

Atomic examples#

Five single-file demos under examples/atomic/. Each one is ~80 lines, covers exactly one framework concept, and runs as a standalone script:

These are the canonical reference for each step in the Implementation Guide.

Elbow flexion#

A complete PID-controlled FES application for elbow flexion. Uses agonist/antagonist muscle pairs (biceps for flexion, triceps for extension), an FFT-based PID auto-tuner, and a PyQt5 GUI with control sliders and real-time plots.

Read it when you need to do continuous joint-angle tracking — the strategy's structure transfers directly to any one-DOF FES application (knee, ankle, wrist).

File: examples/elbow_flexion/

Haptic navigation#

Demonstrates the HapticLibrary extension point. Arrow keys in a GUI fire haptic cues on different parts of the wearer's body — belly, back, shoulders. No closed-loop control logic; this is the canonical example of using EMS as tactile feedback rather than functional movement.

Read it when you need custom haptic patterns, named playable slots, or any non-functional use of EMS hardware.

File: examples/haptic_navigation/

Generic GUI#

A reference PyQt5 GUI template using the framework's reusable widget toolkit (fes_framework.gui.app.FesApp, fes_framework.gui.data_adapter.DataAdapter, plus tab classes for overview and sensor data). Comes with a no-op DummyStrategy so you can see the GUI light up without any real control logic.

Read it when you're starting a new GUI-based application from scratch and want a clean template to copy.

File: examples/generic_gui/

Walking FES#

The full reference application — gait-phase-adaptive walking FES for foot drop / post-stroke gait rehabilitation. Five muscle groups per leg (10 total), gait-phase-adaptive timing, ML step detector, comprehensive PyQt5 GUI with three tabs.

This is the template project. Read it last; it pulls together every framework feature.

File: examples/walking_fes/

How to run any example#

From the repo root, with the Teslasuit connected and Control Center running:

python -m examples.<name>.main
# e.g.
python -m examples.elbow_flexion.main
python -m examples.haptic_navigation.main
python -m examples.generic_gui.main
python -m examples.walking_fes.main

Atomic examples run as direct scripts:

python examples/atomic/minimal_closed_loop.py

Press Ctrl-C to stop. The framework's unconditional cleanup ensures EMS is muted and mocap is stopped before the process exits.

Hardware requirements#

All examples target real Teslasuit hardware. There is no mock / simulator path. If the suit isn't connected, SuitHandler construction will fail during engine initialisation. See Installation for the full hardware setup.