Project Drive

From MAE/ECE 148 - Introduction to Autonomous Vehicles
Jump to navigation Jump to search


Sparkfun AVC Speed Demon Class


IEEE at UC San Diego strives to provide accessible hands-on engineering opportunities to students with initiative. Project Drive is one of our two advanced annual projects that provides a robotics challenge to a small team of experienced students. The team is challenged to design a small autonomous ground vehicle that can quickly navigate through an outdoor terrain course at the annual Sparkfun Autonomous Vehicle Competition in Niwot, CO on October 14-15th. The Speed Demon competition is described online:

Competing vehicles will be expected to complete a course littered with obstacles within a set time frame without human interaction. The track is a figure-8 layout with a major axis length of approximately 178 feet and a minor axis of approximately 78 feet.

Those course obstacles test the team's ability to integrate robotics systems into the vehicle. Since 2006, IEEE at UC San Diego has sponsored Project Drive (as it is currently referred) to create a multidisciplinary engineering experience common in industry. The team of students must learn and apply techniques for hardware and software design, research skills, as well as project leadership. Technical skills applied during this project include: Physical design with CAD Analog circuit design for power regulation Signal processing with sensors such as GPS, LIDAR, ultrasonic, motor encoders, accelerator, gyroscope Control theory with PID Localization/navigation methods using ROS Machine vision using openCV

The entire experience of the IEEE project takes place during one academic year. By Week 5 of Fall quarter the team of 8 students was formed. The remaining weeks of Fall quarter were spent designing a vehicle block diagram, custom LIDAR block diagram, and researching how to work with ROS. WInter quarter was spent ordering and receiving components, assembling electronics onto the vehicle chassis, designing the power system, and fabricating a custom LIDAR system. Spring quarter has been spent tweaking ROS and integrating the controls and localization systems together to create reliable navigation. The team documents the development process every three weeks and plans to showcase the final product on the IEEE at UC San Diego website to inspire robotics in the community.