Howdy everybody! With school starting, I’ll be posting a bit less frequently. But I wanted to share some midweek motivation with everybody. Even though it’s only September, I’ve been thinking nonstop about October and November. Every year for the past three years, my students and I have participated in a hackathon at Virginia Commonwealth University, called HealthHacks. HealthHacks is a hackathon dedicated to solving problems in the healthcare industry. Big shoutout to VCU for hosting the event free of charge and allowing us to jump in!
For those who don’t know, a hackathon is typically a 24 hour sprint to solve a problem, often using some form of coding or computer science. Last year, instead of spending their weekend sipping pumpkin spice lattes and Halloween costume hunting, three students spent it with me, in a mad dash to solve a problem! So, what did we get done in 24 hours? Well, how does a concussion detecting helmet sound?
Backstory: How we Picked Our Project
Our team attended an hour and a half problem pitching session where they listened to doctors, nurses, emergency responders, and others in the healthcare industry explain some of the problems they have in everyday life. We were particularly interested in a problem put forward by a doctor - detecting concussions is a very nebulous process.There is no guaranteed physiological indicator of a concussion, even MRIs and CT Scans cannot say with certainty whether a concussion has occurred.
A concussion usually results not from the initial hit, but from the rapid deceleration of the head. The brain keeps moving and can slam into the skull. Over time, multiple concussions have been shown to drastically reduce the function of the brain and lead to a reduced quality of life. Even more surprising, and tragic, is that the percentage of multiple concussions is highest in high school sports. Without a dedicated and professional trained coaching staff, it is often hard to watch all players and assess individually.
I let the students debate the pros and cons of creating a concussion preventing helmet. I tried to offer some expert knowledge but no opinion. Easier said than done! We decided that we didn’t have the expertise or testing equipment and instead; we wanted to give a way of easily detecting concussions from the sidelines.
Research and Proof of Concept
First, we needed to know what we were looking for. In our research, we discovered that a typical roller coaster puts 5G's (gravities not grams!) on a rider and a human will typically pass out after being under 10G's for longer than a couple of seconds. In professional football, it is not uncommon for a hard hit to exert over 95G's on the player. For reference, a bad car crash can exert 80G's!
We wanted to use an accelerometer in our design of the helmet, but the ones we had could only measure up to 1.5G's, well below the amount we would need to measure to determine if it was a bad hit. Looking back, students reflected this was a positive so that we wouldn't have to give someone an actual concussion just to demonstrate!
For proof of concept, we decided to use a SparkFun force sensor, a handful of LEDs, and an Arduino Uno. Our design uses LEDs on two sides of the helmet so that no matter which way the player is facing, the coach or referee would be able to see the indicator lights. There are two white LEDs on the top of the helmet that show the helmet is taking data. A hard hit of the hand to the sensor will immediately light up red lights to indicate that a player should immediately be checked for a concussion.
However, we didn't think this was enough. Hits below this level could still have a serious cumulative effect. So, a medium hit will turn on a yellow light. Another will turn on a second yellow light. And a third will trigger the red light.
Getting to Business: Coding and Wiring
The rest of the night was spent, as all good hackathons are, chugging coffee, wolfing down pizza and donuts, coding and troubleshooting. (How could it be any better?!).
First, we had to calibrate our force sensor to detect just the right hit that could be reproduced consistently to trigger the lights. Finally, our code worked!
Now we just needed to wire everything up and attach it to the helmet. But where was the electrical tape and soldering iron? Nowhere to be found. So instead we had to individually hand wire all of our lights. The hard part was keeping them from tangling with each other and pulling out of the Arduino pins.
Combine that with everybody coming to our table to ask questions, we were starting to run out of time. Around 11am, we finally had it all finished!
We made a few changes to it based on feedback from doctors and even met a football coach that worked with Virginia Tech and the Eagles. He gave us some pointers about visibility from the sidelines and we made further changes. Our fragile creation was ready. But then, we found out we had to move it to present in a different room - we were terrified it would fall apart in transit! We walked it to the room as if it was made of nitro-glycerin.
Presenting Our Creation to the Judges
My team did a stellar job of walking the doctors, professors, and nurses through our device and answering questions related to it - all in under three minutes. A whole separate post can be made about how difficult it is to sum up all of your research, reasoning for design choices, and demonstrating the product in an engaging way. (Let me know if you’d be interested in that in the comments.) But we saw the value in making a prototype functional but not pretty. The students were so worried about judges seeing the duct tape, and no one commented on it. Definitely a lesson in designing a presentation for others, not for yourself!
So what did we learn? My team kept repeating how they couldn’t believe how quickly the time went. When they first planned out the work, they left out some key steps such as troubleshooting, calibrating the sensor, testing, and wiring the lights. This proved to be what they spent most of their time on.
Even in a small team of three, communication was patchy at first. We started having hourly stand up meetings at my request due to the “tunnel-vision” we all fall victim to. But when you’re working in a team with a timeline, that can be the difference between success and failure.
They also had to learn how to prioritize work based on the skills and knowledge they brought with them. Time management and task planning were the biggest takeaways for them. Two years later, they still are talking about this experience!
For me, as the project manager, I got to watch students grow rapidly in their hard skills, confidence, and critical thinking in a very short amount of time. They performed at the level most businesses wish they could. Students working on a real world problem, without threat of a grade, but still a high risk of failure, led to amazing outcomes. My role was to help guide them, like bumpers on a bowling lane, not plan out the experience for them. Too often we are worried about creating a low risk environment where we have all the answers, but this experience proved to me that students thrive most, outside of comfort.
You’d be surprised at how well people rise to the challenge. Our next HealthHacks is November 3rd this year. I’ll be sure to try and livestream it or keep things updated! If you're in the area, I definitely recommend checking it out.