Browse categories
Explore
Fiverr Pro
English
$
USD
I will build tinyml sensor classification firmware for stm32 and nrf
About this Gig
Want your sensor device to recognize motion, vibration, impacts or other events locallywithout relying on a cloud connection?
I work on embedded sensor acquisition, DSP pipelines and practical TinyML/Edge AI integration for resource-constrained microcontrollers.
What I offer:
IMU, accelerometer and vibration sensor integration using I2C/SPI
Accurate sampling, buffering and timestamping
Filtering, FFT, RMS, spectral features and event-window extraction
Python data preprocessing and labeled-event evaluation
Rule-based baselines before machine learning when appropriate
Compact inference integration on supported STM32/nRF targets
Measurement of CPU load, latency, RAM, flash and power impact
My embedded experience covers Nordic, STM32, Zephyr and real-time high-rate sensor acquisition.
Packages cover a prototype; new PCB design, dataset collection, regulated medical algorithms and cloud dashboards are separate.
Share your sensor model, dataset and required detected events before ordering.
Platform:
STM32
Sensors:
Accelerometer
•
Sound & vibration
•
Gyroscope
Expertise:
Firmware development
•
AI
•
Signal processing
My Portfolio
FAQ
Can you build a model if I have no dataset?
We can start with sensor data logging and an event-labeling plan. Full classification needs representative recordings and labels.
Will inference run directly on the MCU?
That is the aim of Premium if target memory, compute and input features support it. Feasibility is verified before committing to a model architecture.
