Driven by the passion for Robotics, IoT and Machine Learning, we could not resist the urge to combine them into an awesome and useful system!
What it does
We have built several robots and implemented a few AI algorithms, using which one can:
- Solve the Robotics Challenge (the maze)
- Have sensors' data mapped in Unity and the environment visualized on that
- Use our 'Family Robot' to get things from other rooms and benefit from our facial recognition algorithm to get personalized service and be alerted of presence of unknown people
- Get help when you're not feeling well - with our Google Assistant-linked chatbot
How we built it
We used Arduino and SnappyXO kit and a number of sensors (Ultrasonic, Temperature, Humidity, LDR) to create a robot able to solve the maze challenge, as well as explore the environment and interact with it and the members of family. We used Raspberry-Pi and its camera to run Face Detection algorithm that distinguishes between different family members, as well as detects unknown people and activates an alert. Also, we implemented mapping of sensors' data into Unity and its visualization in real time. For the Health Care bot, the app engine, Google Cloud storage, Softheon workshop UI, DialogFlow, Firebase were used - to create a personalized care-receiving experience.
Challenges we ran into
Installing OpenCV (contributors' distribution) on Raspberry Pi took us more than 12 hours, and we were extremely happy to have finished it by Sunday morning. The WiFi module on Arduino refused to work, which we overcame by using serial communication. Being not at all familiar with Unity, we found it very uneasy to get started with it and implement sensors' data visualization. Integrating different outcome format from RESTful interface and premature third-party library was challenging.
Accomplishments that we're proud of
In a short period of time, we built a nice little robot, equipped it with a number of nice tricks and skills, as well as developed visualization of the environment in Unity and a health care bot. We believe to have built something useful, with a mission of improving people's lives.
What we learned
As always, this collaboration with bright and inspiring peers was a great experience for us! We deepened out skills in Robotics, IoT, Machine Learning, and Artificial Intelligence - each contributing to this broad but wholesome project.
What's next for the United Sensors of America
Take over the world!