XiLoop — X-in-the-Loop, from software to hardware

XiLoop wires your controller to a simulated plant, runs the closed loop, and verifies your requirements automatically. Think pytest for control loops.

[ Device under test ] --command--> [ Plant model ]
        ^                               |
        +---------- measurement --------+
              LoopEngine (fixed rate)
   CampaignRunner: testplan.yaml -> PASS/FAIL report

• Device = the thing you're testing — a Python controller today; emulated firmware (Renode/QEMU) and real boards over serial on the roadmap.
• Plant = the simulated system — an actuator, motor, battery, vehicle… write your own in ~15 lines.
• Campaign = a YAML test plan with requirements (rise time, overshoot, steady-state error…). XiLoop runs the scenarios and emits a pass/fail report.

Quick start

pip install -e .[dev]
python -m examples.actuator_pid.run     # run the demo campaign
pytest -q                               # run the test suite

from xiloop import PID, PIDDevice, ActuatorPlant, CampaignRunner

device = PIDDevice(PID(kp=2.0, ki=1.0, kd=0.05))   # the controller under test
plant  = ActuatorPlant()                            # the simulated system
result = CampaignRunner(device, plant).run("examples/actuator_pid/testplan.yaml")
print(result.summary())                             # ... PASS/FAIL per requirement

Interactive tuning (v0.2)

pip install -e .[plot]
python -m examples.actuator_pid.tune   # sliders for kp/ki/kd + setpoint, live metrics

Drag a slider — the loop re-runs instantly and rise time / overshoot / steady-state error / settling time update live. When the response looks right, lock the gains in with a campaign.

Why

In-the-loop testing (SiL/PiL/HiL) is how industry verifies control software before touching real hardware — but the established tools (dSPACE, NI, ECU-TEST) cost five to six figures. Students, makers and small teams rebuild the same plumbing every project. XiLoop is that plumbing, done once, open-source.

How it compares

The pieces of this workflow exist — the combination doesn't:

Existing tool	Great at	What XiLoop adds / does differently
OpenHTF (Google)	Hardware test campaigns, measurements, pass/fail	OpenHTF orchestrates bench tests on physical devices — no closed control loop, no plant model. XiLoop closes the loop against a simulated plant.
python-control, bdsim	Control-systems analysis & block-diagram simulation	Analysis libraries — no requirements-as-YAML, no pass/fail reports, no device-under-test abstraction toward firmware/hardware.
PX4 / ArduPilot SITL	Superb simulation-in-the-loop	Locked to their own flight stacks. XiLoop is domain-neutral: any plant, any controller.
Renode / QEMU	Firmware emulation	They emulate the chip; no plant, no campaigns. XiLoop orchestrates around them (roadmap v0.3).
dSPACE / NI / Speedgoat / ECU-TEST	Certified, industrial HiL	Five-to-six-figure commercial rigs. XiLoop is the free, software-first on-ramp — not a certification replacement.
Hand-rolled scripts	Quick start	Rewritten badly every project. XiLoop is that plumbing, done once.

One line: pytest for control loops.

Roadmap

Version	Adds	Status
v0.1	Core engine, Device/Plant interfaces, telemetry, campaign runner, actuator example	✅
v0.2	Interactive tuning dashboard (sliders, live metrics) + settling-time metric	✅ this release
v0.3	SocketDevice → C/C++ firmware in Renode/QEMU (PiL)	planned
v0.4	SerialDevice → real microcontroller (true HiL), PyPI release	planned

License

MIT — © Muhamad Alif Izzuwan Bin Ibrahim (github.com/Alifizz01)