Two years ago, RoboNav began their long-term plan to move away from the Intelligent Ground Vehicle Competition (IGVC) and instead begin participating in the University Rover Challenge (URC). We previously introduced this upcoming change in the March 2021 newsletter. The University Rover Competition still keeps the core focus of RoboNav, autonomous navigation, while increasing the difficulty and complexity of the mechanical, electrical, and software subsystems onboard the robot. The robot must not only traverse rocky and uneven terrain, but also communicate with a tele-operation station and perform dexterous tasks with a robot arm. The competition also includes a new-to-Robonav science challenge which involves analyzing soil for signs of life.
The team has been preparing for this huge new challenge throughout the last two years. The administrative planning has long been complete, and the team has already made great strides in their technical groundwork. We interviewed the team leads to understand their progress so far and upcoming goals and challenges for this year.
The mechanical subteam, led by Matthew Fernandez, provides the base functionality and mobility for all other teams. Throughout the years in IGVC, the team has learned how important it is to provide a good amount of software and electrical testing time with a working, moving robot. This is even more true when there are many new systems to roll out and test, as is the case this year with the new challenges brought on by URC. Because of this, the mechanical subteam has been working hard to finalize their designs and begin manufacturing as soon as possible. You may remember a feature piece from May 2022 which went through the end effector design for the robotic arm, which will be used to tighten screws, push buttons, and open latches in URC’s equipment servicing mission. The overall arm design has since been mostly completed in CAD.
The drivetrain for the robot also recently went through a design review and received some great feedback thanks to alumni Matt Barulic and Jonathan Spalten. The rover will use a rocker-bogie system to traverse the rough terrain expected in URC. The two rocker-bogie legs are attached to the main chassis with a differential, which allows uneven movement, further stabilizing the chassis of the robot.
Lastly, the wheels of the rover are designed with compliant spokes to allow compression and uniform stiffness. The custom designed wheels will be 3D-printed and attached to the motor assembly with a cage to provide some additional support. The compliance of the wheel design was studied by the mechanical team using FEA to help select the thickness of the material.
The upcoming goals for mechanical this year include training and onboarding their new members, then proceeding with the final design details and manufacturing of the robot.
The software subteam has been hard at work laying the foundation for URC. Though the challenge still includes autonomous navigation, there are many other considerations that come into play with the inclusion of a tele-operation requirement and robotic arm control. Lead Aidan Stickan states that “URC is much different than IGVC because it has a teleoperation component in addition to the fully autonomous navigation mission. Also whereas IGVC was a true black-box SLAM situation on the mapping front, URC allowing satellite images to be parsed in between checkpoints cuts back on the mapping [and] localization complexity significantly.” In sum, the software design for URC focuses more on autonomous decision-making, remote communication, and inverse kinematics, whereas IGVC focused more on localization and computer vision.
So far, the software team has written the lower level nodes of the software stack which face hardware and are used to control the rover, like motor controllers and joystick control. One of the changes from IGVC to URC is the implementation of a traversability costmap. The costmap grid is filled with squares that include the height of the terrain, as seen by the robot’s LiDAR. This information is then used to path efficiently while still avoiding steep drop offs or climbs that may present a challenge for the rover. The team is now very close to having a traversability costmap solution, and an early prototype of the behavior tree (critical for enabling the robot to make choices out in the field) has been completed.
Behavior tree
Electrical has primarily been doing research on the major components needed for the rover. Future work will be divided into a few core areas such as motors, firmware development, and communications (radio, etc.). Electrical this year will heavily overlap with other subteams to give feedback and be well prepared for integration.
The science subteam is a new addition to Robonav. Led by Elijah Vasquez, this team is bravely going where no RoboJackets team has ever gone before: biology. URC includes the requirement for the robot to onboard soil samples and analyze them for signs of life. Therefore, the science subteam has identified some basic chemical tests for life detection and will work closely with the mechanical and electrical sub-teams to implement a soil sampling module on the robot. The next steps include testing out the proposed experiments and determining which one would be the best fit in terms of the mechanical and electrical constraints.
After speaking with each RoboNav team lead, it is clear that tons of hard work and great engineering has already poured into their rover, nicknamed Walli. Many of the challenges lying ahead, such as robotic arm control, soil sampling, and tele-operation of the robot, are intrinsically multidisciplinary and will require excellent communication between the subteams. This year’s project manager, Priyanka Rajan, states that “Cohesion of the team as a whole is also a priority. By continuing educational presentations on major design decisions and having frequent all-hands meetings we hope to strengthen ties and create a stronger sense of community between members of different subteams.” Although qualifying for URC in the team’s first year is a big ask, she remains positive. “This year is definitely going to be a big challenge for RoboNav, first because this is our first year working toward the URC competition, and second because we’re very ambitious in our goals. I’m really excited and optimistic, though, because we’ve got an incredible group of hardworking, passionate, and dedicated members going into this. Our team has already hit the ground running, so I’m confident we’ll create something great. I can’t wait to show everyone what RoboNav can do!”
We will check back in a few months to see even more progress from the Robonav team. Here are some important dates for the students coming up:
- October 26, 2022: URC Registration Deadline
- December 2, 2022: Preliminary Design Review Deadline
- March 3, 2023: System Acceptance Review Deadline
- May 31 – June 3, 2023: URC2023 Finals
If you’re interested in learning more about the URC guidelines and challenges, check out the 2023 official rulebook. If you’re interested in participating in future design reviews, like the one for the drivetrain or robotic arm, join the alumni slack to get notified.