Overview

  Air Vehicles   Ground Vehicles   Autonomous Boats   Data Gathering   Interfaces  
MLB Bat UAV  

Ground Vehicles

GPS Car

The Navigator was developed while Prof. Holden worked at MLB for a middle-school after school program. The idea was to attach an autopilot computer and a GPS to the radio-controlled cars the students built to make the car autonomous.

It was an interesting project using just GPS for navigation, since the GPS has an accuracy of only 20 feet or so and updates postion and velocity only every second. The car works well, and with a few other sensors (bump switches or some obstacle detection would be nice!) the car could be used for farm delivery or other autonomous tasks. It is a great demo platform too, and with the single sensor the autopilot is easier for students to understand.

Movie of the GPS car on the SF State quad, out to a waypoint and back. (3.5MB)

Electronics

The electronics are based on the Atmel Mega 32 processor and programmed in C with the Codevision compiler, using lots of floating point math.

Car Electronics

User Interface

The LCD display and buttons are mounted in the top of the car, and allow the course to either be downloaded from the GPS (set up as a standard GPS route) or entered by carrying the car around the course, or entering the latitude/longitude/speed manually.

The user can also adjust the feedback control gains and calibration constants using the interface, so there is no need to re-program the chip to tune the control law.

User Interface

There is a rough-draft user's manual here.

Control Law

There is some basic information about the control law and estimator in this paper.

Small Autonomous Boat  
GPS Guided Car  
Single Sensor Autonomous Aircraft  
Autonomous Car

 

 

Water Sampling Autonomous Boat  

 

Prof. Holden Contact Info