ECE 110/120 Honors Lab Section : iMASK

Statement of Purpose

With the current pandemic, masks are a necessity in everyday life. If one wants to watch a movie, eat at a restaurant, or study at a library, one needs a mask. However, masks can be easily forgotten or lost when they are needed the most. Our solution to this problem is iMASK, an autonomous vehicle that carries free masks around the campus for those who unfortunately lost or do not have their masks. iMASK can follow a designated path, avoid obstacles, and recognize humans and common hand signs for commands.

Background Research

In fact, autonomous vehicles are in use already in our daily lives. At Purdue University, Purdue Starship, autonomous delivery service, delivers food and drinks to students ordered through an app across campus. Moreover, company Tesla mass produces self driving cars all over the world since 2015. With this technology, one does not have to actually control the vehicle itself, thereby reducing the stress and effort, allowing to focus more on some other tasks that could be more important. Especially in a pandemic, having someone to deliver masks could go against the social distancing guidelines; however, making the vehicle autonomous would mean that this process would be completely contactless. 

Our team tried to come up with multiple ways to integrate self driving to a vehicle that can deliver masks, and concluded that utilizing OpenCV with a Raspberry Pi was the most manageable way to do so. OpenCV is a free library that can be used in Python for image recognition. We also have prior experience with OpenCV, and think it would work well for this project. By using OpenCV with Python, it would also integrate well with a Raspberry Pi which works with Python and can be used to control our vehicle.

To build the vehicle itself, we could use DC motors to control the wheels. We can also use a Servo motor to control the lid of the container in order to deliver masks. These will all be controlled by a motor controller and a Raspberry Pi that will be attached to the body of the vehicle. As input, we will have a camera module that sends information to the Raspberry Pi for image recognition. 

Block Diagram / Flow Chart

System Overview

1. Battery is the main power source 
2. Raspberry Pi is the computer that controls everything
3. Speaker will outputs selected audio files for certain situations
4. Camera will detect visuals such as obstacles hand signs made by people
5. Motor Controller controls motors on the vehicle
6. DC Motors control the wheels 
7. Servo Motor controls the lid

Parts

This is total of $61.74. We are planning to save some of the budget until later of our project in case we need extra parts if some of the parts go missing or break.  

Possible Challenges

We need to build a vehicle that is a reasonable size but also needs to carry masks, Raspberry Pi, speaker, camera and motors. This will be an obvious challenge since we have to work with software while worrying about the physical design aspects of the car itself. We need to make sure that the car can actually before we implement the self driving ability as well as other modifications. Moreover, there will be some delays in the motors and components of our car. Making sure that all of the steps taken by our vehicle is completed within a reasonable time is also important as the mask delivery process should be swift and accessible. Therefore, ensuring that the vehicle detects objects in real time would take us a decent amount to accommodate the trial and error process. Since we do not have much time, we need to come up with some prototype to start testing right away. 

References

[1] "Autopilot", Tesla.com, 2022. [Online]. Available: https://www.tesla.com/autopilot. [Accessed: 26- Feb- 2022]

[2] P. Purdue Marketing & Media, "Purdue University welcomes delivery robots", Purdue.edu, 2022. [Online]. Available: https://www.purdue.edu/newsroom/releases/2019/Q3/purdue-university-welcomes-delivery-robots.html#:~:text=9)%20launched%20robot%20food%20delivery,anywhere%20on%20campus%2C%20within%20minutes. [Accessed: 28- Feb- 2022]

[3] "Self-driving RC Car using Tensorflow and OpenCV — The MagPi magazine", The MagPi magazine, 2022. [Online]. Available: https://magpi.raspberrypi.com/articles/self-driving-rc-car. [Accessed: 28- Feb- 2022]