Edge AI apps

Edge AI apps · Course page

Build an app focused on a single model, wire a verdict to an action through a pipeline that guards against false positives, fuse it with a raw sensor, and publish events over MQTT.

Module goal

Turn a model’s answer into a trustworthy action, and send it off the board with discipline.

Lessons

Lesson	Topic	Time (min)	Slides
6.1	Six models and the edge_ai API: an app focused on one model	70	slides.md
6.2	Hands-on: our own focused app	75	slides.md
6.3	An action pipeline: CONF_FLOOR, debounce, cooldown and on_result	70	slides.md
6.4	Hands-on: an action pipeline that guards against false positives	75	slides.md
6.5	Sensor fusion: a model’s verdict with a raw sensor	70	slides.md
6.6	Hands-on: publishing a fused event to MQTT	75	slides.md

Lessons come in pairs: a concept lesson followed by a hands-on lesson with a practice file, a solution, and a lab.

Module checkpoint

You pass this module once you can do all of the following (details are in the Lab section of each hands-on lesson):

• A single-model focused app with a clean UI, targeting a model with find_model(), showing the verdict, every class’s bar, and latency, with a counter that genuinely fires when the target class crosses CONF_FLOOR, retargetable to at least two models (lesson 6.2).
• A debounced action pipeline where the target class continuing triggers a real action, while a brief flicker of the signal is guarded against (lesson 6.4).
• A fused decision (verdict AND a raw gate) genuinely publishes to MQTT, once per event, while gentle motion is never sent (lesson 6.6).