-
Notifications
You must be signed in to change notification settings - Fork 2
Home
Using the PYNQ-Z1 board, our goal is to create an autonomous obstacle-avoiding robot that traverses around a designated area, such as a living room, and maps out the relative location of all obstacles--such as walls, couches, or even a PlayStation. This is done with an ultrasonic sensor and a servo motor, rotating and scanning its surroundings and storing what it senses in its memory. For example, in a two-dimensional array of 10 units by 10 units, the robot would stop in the center of one of any 100 square units to scan left to right for a full 180˚ of range and marks down any objects that it sees within 0.5 units to √2 units of range. This robot would then move onto an open, adjacent unit-square to repeat the process again (scan area, locate objects, and move on) until all 100 unit-squares are mapped.
Week 6: Because of the lack of pynq support, we are forced to find workarounds for the HC-SR04 and servo peripherals. Our goal for the end of this week is to have all of the sensors working so we can start building and the device.
Week 7: Our goal for the seventh week is to work out and implement a way for our servo and HC-SR04 sensor to work in tandem to scan an area around the device and store the data into an array on the PYNQ board.
Week 8: Attach wheels and car frame to the PYNQ device and make it mobile. This will require us to rework our mapping algorithm to include the change in position associated with the newly acquired movement.
Week 9: Continue to implement the changes in algorithms that are dependent on the movement of the car as well as debug things that are wrong with the device.
Week 10: Debug and make the final touches on the way the device moves as well as work on the documentation and website for our presentation.
For the first few weeks of this course, we spent most of our time researching and looking into the various ways in which we could utilize the PYNQ-Z1 board to create a collision-avoidance system for autonomous vehicles. Using computer vision was one of our original ideas; however, due to time constraints and the small team size, we chose to use an ultrasonic sensor to help determine the distance to an object or obstacle that it detects while scanning. However, because we realized our projected seemed a bit too ambitious (with Kasra’s words of advice), we will potentially be simplifying the project to program the PYNQ board to store key value pairs of angle and distance, using a stationary robot that scans an area left to right.
- HC- SR04 Ultrasonic sensor x2
- Servo Motor x2
- Arduino Car kit including:
- 2 DC motors
- Car frame
- Battery pack
- Tomohiro Ohkubo (CSE) - Software and Documentation Specialist.
- William Floyd (EE) - Sensor Implementation for HC-SR04 and Servo Motor