iOS Algorithm Engineer (CoreMotion / Physics)

We are a sports technology startup building an Apple Watch app for sports training. We have a functioning app that records raw accelerometer and gyroscope data. We need a specialist in Signal Processing and CoreMotion to help us tune the algorithm.

Engagement Type: Short-Term Contract / Moonlighting Friendly

• Time Commitment: Estimated ~10–15 hours total.
• Schedule: Flexible. This work can be completed on evenings or weekends. We focus on the deliverable (the algorithm), not the hours.
• Location: 100% Remote.

The Mission: We are tracking a specific arm movement (throws). We have the raw data logs (CSV). Your job is to analyze this data and provide the logic/code to:

1. Accurately count "Reps" (Throws).
2. Distinguish between a real throw and "Noise" (e.g., dropping the arm, walking, picking up a ball).
3. Handle the drift/noise inherent in MEMS sensors.

What We Have:

• Existing WatchOS app that logs data (User Acceleration, Rotation Rate, Gravity) at 50-100Hz.
• Dataset of "Clean Throws" vs. "Noise" (CSV format).
• A lead iOS engineer (Swift) ready to implement your logic.

What We Need From You:

• The Algorithm: You don't need to build the UI. We need the mathematical logic (or Python/Swift function) that takes the raw inputs and returns a "Success" count.
• Drift Correction: Experience using Complementary Filters, Kalman Filters, or Sensor Fusion to clean up the data.

Requirements:

• Deep experience with Apple's CoreMotion framework.
• Strong background in Vector Math and Physics.
• Experience with Human Activity Recognition (HAR) or Gesture Recognition.
• Ability to work fast

This project focuses on generic signal processing math (physics/vectors) and does not require using proprietary IP from your current employer. We are happy to sign a customized IP agreement to ensure you are protected.

• Skills

  • Algorithm Development
  • watchOS
  • Swift (Programming Language)
  • Core ML
  • Signal Processing
  • Mathematics