Tracking Team:
Evaluating Criteria:
The probability that the algorithm can accurately detect a person must be high for all of the use case distances. (The algorithm puts out a confidence probability). This will be used to analyze the information collected. If the predicted distance compared to the actual distance away from the camera is within a to-be-determined reasonable margin of error. If the algorithm currently determines and keeps a person as the “desired person” even after several other people come into and out of frame at different distances, sizes, and colors, then the team will consider this prototype a success, and the team will be ready to continue to next steps and integrate with other sub-teams.
Results:
The Tracking Group found that the program was able to identify a person and follow them. However, when the desired individual was blocked from the camera by another the tracking was lost. A recording of a trial can be seen here.
Electrical Team:
Evaluating Criteria:
The robot must provide light for the user to be able to see to move and for the camera to be able to see the objects it is tracking. This meant choosing headlights bright enough to cast enough light to see, but not to blind any other passing people. The alarm light and siren need to be bright and loud enough to bring attention to any ongoing situation. The goal for the alarm light is to be blue to associate with the blue light system. These devices need to be able to be controlled by user inputs.
Results:
The electrical team used the functional requirements for lighting and sound to make sure the selected components would provide enough light and sound. Each of these components have a specified voltage and wattage that they need to operate correctly. These values were used to determine the voltage that the batteries would need to provide. Using voltage dividers the circuit was designed to provide the proper power to each component. This circuit was tested in a SPICE simulation deck to guarantee that each component was getting the proper values. An Arduino with relays was chosen to control when different lighting elements are operational.
Structure Team:
Evaluating Criteria:
In order to evaluate the prototype’s effectiveness, the team will develop multiple tests for Lafayette College students outside of the engineering department to complete. Control panel tests will include timing students to see how quickly they are able to interact with certain buttons. The team will also spend time evaluating options for the wheel size, number of wheels, and wheel layout based on physical campus obstacles. Finally, the appearance team conducted student testing on colors and shapes. The results of these tests help to determine what emotions different colors or shapes evoked in students.
Results:
The appearance team determined that curved shapes and blue-purple colors are the most calming. The wheel base team chose a 6 wheel bogie chassis. This allows for the robot to climb a standard 8 inch curb. Finally, the button panel chose a final design that allowed for quick and easy use of the device.