Python and IoT: From Chips and Bits to Data Science
summary
End-to-end project that builds a smart lighting system: sensor nodes running esp8266 with MicroPython publish minute samples over MQTT to a Raspberry Pi which stores ~25,000 samples per room over six weeks. Data cleaning uses NumPy/Pandas in Jupyter, k‑means to discretize states, then per‑zone hidden Markov models via hmmlearn (about 200 on/off transitions per room) and a lightweight event player that sends transition events to Philips Hue via REST. Runtime glue and deployment use a custom ThingFlow framework.
AI-written summary · about
chapters (14)
- 0:00Python and IoT project intro
- 1:04Why Python for IoT
- 3:13IoT hardware options
- 9:06Building IoT sensor hardware
- 12:34Data capture with MQTT
- 17:35Raspberry Pi setup
- 18:19Analyzing noisy IoT data
- 20:25Preprocessing light sensor data
- 21:50Machine learning for light patterns
- 24:35Hidden Markov models
- 27:26Training by time of day
- 29:55Replaying lights with Hue API
- 32:26Project lessons learned
- 33:36Q&A
related talks
32mData Science in the Internet of Things with Python and Spark
32mIt's dark and my lights aren't working (an asyncio success story)
33mSmart and Simple Data Driven Hydroponic Sensing Applications Using Python, Flask & Raspberry Pi
39mBuilding smart IoT applications with Python and Spark
57mPython for the Internet of Things
46mProgramming the Web of Things with Python and MicroPython
31mProgram all the things - How to develop IoT devices using MicroPython
28m"Build Your Own IoT! Fun With Python and Raspberry Pi."