Documents

Brainstorming

Very early brainstorming for project direction in Fall 2025. We’re already considering what blend of sensors and markers to use to achieve medical-level accuracy.
Whiteboard from late in Spring 2026. Brainstorming for final implementation details before SVD, and logistics of the SVD presentation.

Electrical

Circuit Schematic

Board Layout

Data Sheets

L298 Compact Solarbotics Motor Driver

SPG30 Brushed DC Geared Motor

DC Motor User Manual

Software Charts

Design flowchart for the point cloud segmentation part of our registration process. It names SAM2, but we would ultimately use a YOLO model for classification instead.

Mechanical

Our custom end effector mount design. It incorporates room for a swappable effector (drill, magnet point, etc), depth camera, and a special marker array in front.

This end effector mount was designed to be 3D printed, then mounted to the end of our KUKA arm with the four screw holes on top. Individual IR markers can be easily screwed onto the array, and an end effector tip of our choice and the depth camera can be secured inside.

Experiments

The main subsystem we needed an offline experiment for was the AR digital twin. Since direct metrics of round-trip network packet latency was not reflective of true motion-to-motion display latency, we performed this experiment:

  • Simultaneously record video of the real world and video from the Apple Vision Pro during system operation, with a visible digital twin.
  • Physically move the tracked bones, and observe their movements in the digital twin.
  • Line up the two video feeds by a consistent signal (system time, the same objects visible in both feeds).
  • Count video frames between real motion and when the digital twin replicates that motion.

Experiment results: Average of 388 ms of motion-to-motion latency. This meets our requirement goal of < 500 ms latency.

Preliminary Design Review

AutoKNEE PDR

Conceptual Design Review

AutoKNEE CoDR

Critical Design Review

AutoKNEE CDR Slides

AutoKNEE CDR Report

Progress Review Slides

PR1: PR1 Slides

PR2: PR2 Slides

PR3: PR3 Slides

PR4: PR4 Slides

ILRs

Team B ILRs Folder

Software

Github Link