A UCSC Chip Design Capstone project By: Ananya Manduva, Jackson Friday, Nathan Nakamoto, Nithin Duvvuru, Rishi Govindan, and Shane Stearns
Our goal for the project is to create a custom ASIC SoC that intakes Accelerometer and PPG sensor data, processes, and creates features for a lightweight, three layer MLP model which determines whether the user is in an good stage to wake up (NREM/light sleep) or a bad time to wake (REM/deep sleep). This output is given to a small PicoRV32 core on the chip which then sends out a GPIO alarm signal to prompt wake up. The chip is set to sleep until a wachdog timer, set prior, fires off, indicating it's time to start checking sleep states.
- src
- All the sources required for build
- cocotb
- Sim sources for testing
- scripts
- WIP: Additional scripts for generating ML model, Sensor data, etc...
- librelane
- LibreLane constraint files
- ip
- PDK and logo
- misc
- Need to move around stuff
Walch, Olivia. "Motion and heart rate from a wrist-worn wearable and labeled sleep from polysomnography" (version 1.0.0). PhysioNet (2019). RRID:SCR_007345. https://doi.org/10.13026/hmhs-py35