The Biomech Payload
The stage the project was aimed at from the start. An arm that mirrors my posture from wearable IMUs and grips on the activity of my own forearm muscles, with the isolation drawn and checked before anything is connected to me.
What this teaches me
- Done when
- An arm driven by muscle activity.
- Progress
- 0 of 6 sessions0%
Sessions 46 to 51
The biomech payload
The stage the project was aimed at: an arm that mirrors your posture and grips from your own forearm muscles.
- Not built yet
Dataset work, no hardware
Implement the five standard time-domain EMG features from their definitions, and report within-subject and cross-subject accuracy separately.
- Not built yet
Your first real signal, safely
Draw and check your own isolation topology before connecting yourself to anything, and recognise mains hum and motion artifact from the trace.
- Not built yet
Threshold control
Drive a continuous interface from a biosignal envelope, and explain from your own fatigue plot why a fixed threshold is not enough.
- Not built yet
Your own classifier, on your own signal
Explain why a model trained today fails tomorrow, quantify it as a cross-session gap, and describe the engineering that narrows it.
- Not built yet
Wearable IMUs and your own sensor fusion
Write a complementary filter and say why it works as a high-pass plus a low-pass summing to unity, then say what a Kalman filter estimates instead.
- Not built yet
Mirroring
Close a loop from a human limb to a machine limb and quote its end-to-end latency against a stated budget.