Our beta prototype is a robot that has safety capabilities and can follow people when commanded. The safety capabilities include flood lighting, illuminating the surroundings of the user with blue light LEDs,  and an emergency state. In the emergency state, the robot will flash a blue light and sound an alarm to bring attention to the user.

The robot will have a panel of buttons including an on/off switch, a start/stop follow switch and an emergency state button.  Once the on/off switch is flipped, the robot will begin to power up and the button functions will be available.  When the start/stop button is pressed, the robot will track people using a webcam and tensor flow lite object detection model. The tensor flow lite software will be modified to detect and track a singular person. Their distance from the robot will be detected based on the size of the person within the camera’s frame. A control system will then use these inputs to control the motors to autonomously follow the tracked person. The webcam, motors, tensor flow lite tracking, and control system will all be controlled and powered by a Raspberry Pi and batteries. If the emergency state button is pressed, flashing blue lights will be triggered and an alarm will start to sound to alert nearby people. These systems will be powered by a 12-volt battery and controlled by an Arduino Mega and metal-oxide-semiconductor field-effect transistor, or MOSFET. 

This robot will have a wheelbase since it is a 6-wheeled robot.  The wheelbase between each wheel will be 1”.  The wheels are driven by the a motor which used a gear to power the central wheel axis. Then a chain system was set to power the front and rear wheels from the central wheel axis. This made the robot 6-wheel drive, which allowed for easier turning as well as it improved grip on various uneven surfaces to easily track the user. One capability that expected, however was not achieved during the testing phase of our prototype, was the clearance of the robot could not go over curbs. This was one thing we would like to explore more upon in the future, where improvements will be made to achieve this. 

The audience for our prototype are Lafayette students who wish to be walked home at night and people involved in emergency situations. The functional requirements that will be evaluated are remain with the user, make the user feel safer, illuminate surroundings, and bring attention to the user.

The main method of evaluation will be field testing and stakeholder feedback through surveys. As all of the functional requirements will need to be evaluated, different tests will need to be created. For the ‘Illuminate Surroundings’ and ‘Brings Attention to User’ requirements, the robot will be placed in an open area and people will be positioned at varying distances from the robot. The robot will then turn on the lights and alarms and physical observations will be recorded from volunteers and collated to find common feedback. For the remaining functional requirements, the robot will be programmed to follow a person in various situations. Similar to previous tests the volunteer will provide their feedback and physical observations will be recorded. In addition, tracking confidence and motor performance will be monitored in case there are any substantial issues noted during operation. Finally, feedback regarding “Does the robot make you feel safer,” as well as “Will you use the robot,” was collected. 

For the design attributes listed previously, the evaluation will be similar to the functional requirement testing. Terrain testing will be done at the same time as the following capabilities. This will be done by forcing the robot to navigate unfavorable terrain such as cracked sidewalks, street curbs, grassy areas, hills, and paved paths. Run time and charging will be evaluated by turning on the robot and timing how long it takes for the battery to drop below a certain point. That data will then be interpolated to simulate the time necessary for the entire battery to be used. Once all of the necessary data has been collected, the results will inform updates that will be made when iterating designs. 

Stakeholder involvement will consist of selecting volunteers for testing the functional requirements and design attributes. This will be an important aspect of the evaluation since each stakeholder will have different views on how the prototype may perform. This will give the team a range of opinions from people with different understandings of the project.