Skip to content
MCMASTEREXOSKELETON

2026–27 team / Software

Turn intent into movement.

The software division has two subteams: Embedded & Controls and AI & Machine Learning. Embedded & Controls develops firmware, communication and control systems. AI & ML develops predictive models, data pipelines and the experiments needed to evaluate assistance algorithms.
Explore the subteams ← Back to the design
McMaster students working together on the exoskeleton.
From the workshop to the competition floor

01 / Current member roles

Inside software.

Responsibilities and expectations from the team’s 2026/27 application. Choose the area that fits your interests and experience.
01 / General member

Embedded & Controls

Build the firmware architecture connecting physical hardware to control logic. Work spans microcontroller code, communication, control algorithms and hardware validation.

  • C / C++
  • STM32
  • Python
  • CAN · I²C · SPI · DMA
  • ROS / ROS 2
  • Git

What you’ll work on

  • Firmware and communication

    Develop C/C++ firmware for microcontrollers such as STM32. Implement CAN, I²C, SPI and DMA paths between sensors, controllers and actuators.

  • Controls and simulation

    Use ROS / ROS 2 to write, simulate and tune algorithms for exoskeleton kinematics and actuation.

  • Testing and telemetry

    Use Python to build automated test suites, hardware-validation scripts and telemetry logging tools.

  • Integration and code quality

    Work with Electrical to integrate sensors, motor drivers and power systems. Maintain tested, modular firmware with disciplined Git version control.

02 / General member

AI & Machine Learning

Develop the predictive models that interpret pilot movement and inform torque assistance. The work connects research, multi-sensor data and real-time deployment.

  • PyTorch
  • TensorFlow
  • Data pipelines
  • Model benchmarking
  • Edge deployment
  • Visualisation

What you’ll work on

  • Research and model development

    Review literature on biomechanical intent and torque prediction. Adapt architectures and train models using PyTorch, TensorFlow or similar frameworks.

  • Data and visualisation

    Process multi-sensor streams and extract predictive features. Build dashboards for model metrics, training runs and live telemetry.

  • Benchmarking and validation

    Run controlled experiments comparing accuracy, latency and generalisation against established baselines.

  • Deployment and reproducibility

    Collaborate with Embedded & Controls on edge deployment and hardware-in-the-loop testing. Maintain version control, clear documentation and repeatable research workflows.

02 / Connecting the disciplines

From movement to assistance.

Sensing, prediction and control connect the wearer’s movement to the powered joints. Each part of the software works through interfaces with the suit’s physical hardware.

Software / System flow

Select a stage to explore

Sense. Predict. Control. Assist.

Software follows a real information path: it reads movement, estimates intent, calculates commands and carries those commands to the powered joints.

Inside the flow / 01

Sense movement

Body-mounted sensors provide measurements of how the wearer moves. Software reads and organises those measurements so they can be used for prediction and control.

Sensor integration · Data acquisition
Explore the suit

03 / Take the next step

Tell us where you want to contribute.

Review the responsibilities and tools for your preferred subteam. The application lets you share your interests, skills and projects.

Explore the engineering divisions

01

Mechanical

02

Electrical