Core Members will be placed in Groups based on their skills, experience, and interests. At this time all applications are being accepted on a rolling basis.
Fall 2021 - COVID-19 Information
WATonomous will be operating both remote and in-person this Fall 2021. Meetings are held remotely. We will still have limited physical access to our vehicle for testing and development if required.
The Software Team has invested a lot of time in Spring 2020 building a full software-in-the-loop testing system for our entire autonomous stack using CARLA and Mathworks tools. As such, remote development will be very simple.
The Perception Group is responsible for processing camera, LIDAR and radar sensor data to identify objects of interest (eg. road lines, traffic signals/signs, obstacles etc). Members will research and develop object detection and classification algorithms, experiment with existing open-source libraries, and work with other teams to integrate the Perception module into the software system. Members will gain experience working with machine learning, computer vision, point cloud data processing, system integration, ROS, OpenCV, C++, and Python. They have three current goals:
Migrating our neural networks to run inference on GPU
3D Object Detection
Improving upon our current frustum clustering approach through incorporating state of the art methods
Sim2real Lane Detection and Classification
Conducting unsupervised domain adaptation from simulation to real via adversarial generative method
Overall goals for 2021-2022: Have a vehicle that can overtake others and react to their actions. Also, publish papers and contribute back to the automated driving software community.
Leveraging the ROAD Dataset, explore state of the art solutions to action classification in road scenes
Implement novel action detection architectures to improve on the benchmarks from the AVA and ROAD Datasets
Deploy the network into the WATO automated driving stack
Environment Modeling and Planning
A software engineering heavy project, here we aim to contribute back to the open-source mapping community built around Lanelet2 with a design for integrating online observations with a pre-built map
Expand the feature set of the model to incorporate other vehicles and their predicted actions from the ROAD detector
Kinematic and tire force modeling fidelity assessments in CarSim and on-track
Static and obstacle constraint formulations for non-linear program solvers
Traffic rule constraint formulation, e.g. road boundaries and stop commands from behavioural planner
Rowan Dempster is the Director of Path Planning for 2021-2022. If you have any questions about the projects or your application, feel free to email him at firstname.lastname@example.org.
Infrastructure & Simulation
WATonomous proudly operates our own compute server cluster so that our members have access to top-of-the-line hardware to develop state-of-the-art algorithms. The subteam has two objectives: develop tools for and maintain the WATO server cluster plus developing simulation integration tools.
Building top of the line servers and provisioning Linux and windows-based virtual machines on top of type 1 hypervisors using Proxmox
Hosting distributed network storage for fast access of autonomous driving datasets for deep learning
Setting up and managing network configs for cluster networking and interfacing between the WATO intranet, the UW intranet, and the outside internet.
Building web apps for VM status reporting and management.
Minimum requirements: experience with Linux administration
Nice-to-haves: networking (e.g. DHCP & DNS maintenance), file systems, QEMU/KVM, Docker, Python scripting, Gitlab / Github CI/CD pipelines, full-stack development
Building scalable and reproducible real-world simulation environments for autonomous driving
Creating complex high-accuracy vehicle models using CarSim, Simulink, and CARLA
Using UE4 to facilitate simulation deep learning for real-world autonomous driving
Nice-to-haves: Docker, NVIDIA GPUs, ROS Python / C++, vehicle dynamics, control theory, Unreal Engine 4, Unity
Validate those stacks in both simulation and test track in real life!
Nice-to-haves: Docker, NVIDIA GPUs, ROS 1/2, Unreal Engine 4, Unity
Power Systems Advisor
WATonomous is looking for an experienced person in electrical engineering to assist in designing custom traffic lights, maintaining the power systems of our car, and general electrical tasks. This role is on-call/advisory where you would help the team to troubleshoot any electrical issues, perform any electrical routing and assembly for new sensors, and provide guidance about the vehicle’s power systems.
Minimum qualifications: Experience in electrical assembly and cabling, experience in electrical safety and working with high voltage equipment, and availability to travel to Waterloo on occasions.
Preferred qualifications: Knowledge of power systems for electric vehicles and experience working with computer and sensor hardware.
The Sensor Interfacing group is responsible for calibrating and performing general maintenance on the sensors, and improving the car’s sensor suite. In this role, you will have the opportunity to work with industry-standard sensors and gain a better understanding of autonomous vehicles. For the Winter term of 2022, some projects the Sensor Interfacing group will work on are:
Researching sensor operation in various weather conditions
Developing an auto-calibration algorithm for cameras and LiDARs
No prior technical experience is required as you will have many opportunities to learn but a willingness to learn is required.
Minimum qualifications: Proficiency in Python/C++ and knowledge of Git/version control
Preferred qualifications: Knowledge of ROS, Linux, and basic networking. Experience with robotics, sensor hardware, and/or autonomous vehicles.
RADAR Clustering & Tracking
RADAR is a critical data source for autonomous vehicles as they provide very accurate relative velocity calculations - something cameras and LiDAR struggle with. A strong initiative over the last couple of years has been the development of RADAR for WATonomous. We are looking for dedicated students to help build a sensor-fusion module for our RADARs. RADARs produce a pointcloud which then needs to be clustered into objects, filtered for abnormal detections, then fused in real-time with our LIDAR and camera detections.
You will be working closely with the perception object tracking team and the RADAR team as we finalize the requirements for the RADAR detections and integrate the detections into our existing Kalman Filter object tracker.
Minimum qualifications: Proficiency in Python/C++ and knowledge of Git/version control
Preferred qualifications: Experience with MATLAB / Simulink, ROS, CARLA simulator, sensor hardware, robotics, and/or autonomous vehicles. Knowledge of pointcloud clustering, classical computer vision, and object tracking.
The Sensor Mounting group is responsible to mount all sensors safely, and design necessary enclosures for these sensors. The group performs analysis using tools such as ANSYS to verify design choices and to ensure all sensors are stable during motion. These sensors include cameras, LiDARs, and GPS. The group works closely with electrical and software teams to determine what sensors are required, how they need to be mounted, and enclosure requirements.
Infrastructure & Compute Rack
The group is responsible for outfitting our Test Track with the necessary equipment to simulate competition conditions as closely as possible. This allows for more holistic testing of the car’s performance. Recent projects include designing and building a mobile rig for a pedestrian dummy and designing a mount for traffic lights.
The group is also responsible for maintaining and improving upon the current Compute Rack in the car. The group ensures that any new equipment mounted onto the car is connected effectively to the compute rack. Furthermore, the group directly deals with any possible improvements to the Compute Rack design.
Sponsorship & Outreach
The Sponsorship & Outreach Group is responsible for maintaining both internal and external sponsorships, as well as forging new relationships for WATonomous. This group is our main source of funding, generating over $25,000 each year. We are looking for individuals with high levels of initiative to find new opportunities for our team, and who are capable of writing and pitching proposals.
The Marketing Group is in charge of engaging external parties such as students, faculty members, and sponsors on the team’s progress. The Group is responsible for building WATonomous as a brand. We're currently looking for someone with an interest in videography to organize a publicity video highlighting our progress in the last year on our autonomous stack. You'll be talking with and interviewing members across all divisions and directing the vision for the project.
The Graphic Design Group is responsible for creating and updating designs for WATonomous. This role involves creating attractive and user-friendly website layouts, promotional materials, swag, and car wraps. You’ll also have the opportunity to help set our own brand guidelines. We are looking for someone who loves working with Photoshop, Illustrator, InDesign, or Sketch!
WATonomous is looking for its first web designer to help build out a new website for our team. Our team needs someone to help redesign our current website such that it is easily maintainable, cheap to host, and responsive. We are looking for someone who has an eye for design and is able to create a website that is user-friendly, effective, and visually appealing.
Skills: HTML/CSS/JS, front-end framework such as React, experience building static web pages, hosting
For resources regarding all positions, see our resources page