WEEK TWO
The second week was dedicated in building and testing for the first prototype for the sensor mount. The part was designed using SolidWorks software and was printed on the 3D printer in the lab; this week was the testing phase for the sensor prototype. I worked with the Atmel XMEGA32A4 micro controller that was configured to read analog inputs of the sensor. The micro-controller is part of an existing board for a KnobBot robot that was modified for my testing; the existing potentiometers were replaced with sensor inputs.
Design considerations for the sensor mount:
1. The sensors need to cover a substantial area in order to warn the robot in time for it to take action and stop.
2. The angles for each sensor should be adjusted so as to maximize the range on the sensors.
3. The height above the ground for the sensor mount should lie within the range.
Design considerations for the sensor mount:
1. The sensors need to cover a substantial area in order to warn the robot in time for it to take action and stop.
2. The angles for each sensor should be adjusted so as to maximize the range on the sensors.
3. The height above the ground for the sensor mount should lie within the range.
A linear mount was designed to hold five sensors: the centre one and the two on either end were angled at 37 degrees, and the other two were angled at 40 degrees. The sensors on either end were angled in a way, so as to focus on the line of path or each wheel of the robot. The following figure illustrates the design.
The centre beam is angled so as to identify any positive obstacles directly in front of the robot. The ones on either side of the centre are designed for warnings in case of a drop off or stairs. The ones at either end are designed so as to confirm the warnings and stop the robot.
The centre beam is angled so as to identify any positive obstacles directly in front of the robot. The ones on either side of the centre are designed for warnings in case of a drop off or stairs. The ones at either end are designed so as to confirm the warnings and stop the robot.
This test bench was set up with the mount to simulate stairs and cliffs. The mount is fixed at 27cm above the ground and the readings were recorded with the help of the LCD screen attached to this set up. The LCD would display the change in voltage with respect to the change in distance for the sensor.
Problems encountered during this testing phase:
1. Angles of the sensors: The angles were increased to modify the range on the sensors.
2. The initial testing with the micro-controller did not work too well; the sensor values were inconsistent with the expected values. The analog voltage range for the sensor reduced to 0.1 - 0.004 V which was extremely low.
3. The set up did not work too well in brightly lit room.
The board was replaced with an Arduino Mega 2560 board to check for results and this produced positive results. The voltage range was found to be 2.6V - 0.03V which was closer to the specification.
The design was the first prototype and was modified depending upon the test results and a new prototype was built on the 3D printer. This prototype will be tested next week and examined for further changes.
Problems encountered during this testing phase:
1. Angles of the sensors: The angles were increased to modify the range on the sensors.
2. The initial testing with the micro-controller did not work too well; the sensor values were inconsistent with the expected values. The analog voltage range for the sensor reduced to 0.1 - 0.004 V which was extremely low.
3. The set up did not work too well in brightly lit room.
The board was replaced with an Arduino Mega 2560 board to check for results and this produced positive results. The voltage range was found to be 2.6V - 0.03V which was closer to the specification.
The design was the first prototype and was modified depending upon the test results and a new prototype was built on the 3D printer. This prototype will be tested next week and examined for further changes.