This Hack the North, our goal was to make a wearable hat capable of sensing things around the person wearing it. We were heavily inspired by the concept of spidey sense as seen on the big screen. However, we also felt that we should also build something that society could benefit from.
However, we wanted something more!
We wanted something that has real life applications, and use our technical knowledge to make the world a better place. Thus we came up with Insight4 where we can use our technical knowledge to make people with visual disabilities feel safer.
What it does
Insight4 is a hat with several ultrasonic sensors that can sense objects in 360 degrees field of view, warning the user when an object gets too close, or when an object is moving towards the user with a dangerous speed. We are currently simulating the alarm system with several sound buzzers and LEDs.
How we built it
The project is currently powered by an Arduino leonardo which consists of 4 ultrasonic sensors rotated by a servo. Throughout the event, we designed simulations, drawings, and computed complex mathematical computations for the wearable hat so that it is able to accurately detect any object in 360 degrees field of view. By going through these time-intensive processes, we were able to handle many edge cases in our hat ensuring that the hat followed a stable mechanical and software mechanism.
Once we had a concrete plan, we gathered the hardware components and started experimenting with them prior to assembling them altogether. We then equally allocated the work load where some team members worked on the wiring, 3D printing, Laser Cutting, or the software implementation via C++, and Arduino.
Challenges we ran into
In order to fulfill the design requirements, we decided to use common features in C++ programs such as threads, stacks, and dynamic memory allocation. We were lucky enough to find a StackArray library available but unfortunately, due to the nature of Arduino, multithreading, and dynamic memory allocations were not able to be integrated conveniently. We’ve had instances where the dynamically allocated memory would not get freed properly and block up the heap. In the end we were able to resolve all the problems through research, trial and error as well as the help from the amazing mentors at Hack the North.
Hardware Implementation: When we were testing the hardware and reading the data coming from the sensors, we found that the sensors were not as accurate as we would like them to be. For the ultrasonic sensors, it took a while for the sensor to read the correct values. This made it hard to correctly calculate the “dangerous velocity” of objects coming toward the user.
If you made it this far, please please please note that if you ever 3d print an object with low tolerances, leave AT LEAST TWO MILLIMETRES OF SPARE SPACE SO YOU WILL NOT HAVE TO SPEND AN HOUR FILING YOUR PRINT DOWN!!! >:(
Accomplishments that we're proud of
We are proud of how much we learned during this hackathon. All of our team members did not have much exposure to hardware hacks so it was a great learning experience to be thrown into this challenge. We learned a lot about the hardware/firmware design process and how to debug/fix flaws in our system. Ultimately, we enjoyed this new experience and are looking forward to working on more hardware related systems in the future.
What we learned
During this hackathon, we gained lots of experience working with sensors and coding firmware. We got to expand our experience with working with stacks and managing memory allocation.
What's next for Insight4?
If we were to improve Insight4 in the future, we would add features to improve the user experience. Right now, we simulate the input of the sensors with buzzers and LEDs. Of course for a visually impaired user, this would not be practical. We would change the output to vibration devices, and put multiple vibration devices around the brim of the hat. That way the user will be able to silently know which direction a dangerous object is coming from. Besides improving usability we also would to improve the underlying hardware in order to get more accurate readings.
Log in or sign up for Devpost to join the conversation.