I often go to class hungry, missing breakfast and other meals, because I'm running late or the lines are long and I don't have time to grab food before going to class. I often don't want to miss class material by stepping out to grab food.
What it does
It's a robot that navigates the campus autonomously (or via teleoperation) using a variety of sensors such as LIDAR, ultrasonic, infrared and visible light cameras and process the surroundings using computer vision. Indoor and outdoor navigation is realized using open-source Simultaneous Localization and Mapping (SLAM) algorithms and applications.
Students, faculty, and staff can order food from their local restaurants and dining halls via apps such as Tapingo or Seamless (possibly an in-house app as well) and can track their food as the robot receives the prepared order from dining hall/restaurant staff and delivers it directly to the classroom, office, or dormitory on campus.
How I built it
Using an ODROID-XU4 on-board computer running Ubuntu 15.04 and ROS connected to a modified scooter chair for its motor, chassis, and battery. A web interface for remote teleoperation is accessible over the Internet. Live video streaming from an HD camera will also be piped into the web interface. Autonomous navigation is currently in development.
Challenges I ran into
Open-source packages that don't install easily or don't work properly. I'm looking at you, WiringPi-PHP. Having to downgrade ROS and reinstall from source because certain packages don't compile properly on the ARM-based ODROID-XU4.
Accomplishments that I'm proud of
Being able to remote control the robot over the Internet.
What I learned
ROS, WiringPi, ODROID platform, automated reverse SSH tunneling
What's next for Arova
Autonomous navigation is currently in development. Everything else needs to work.