top of page

Contact

Photo Dec 09 2024, 2 50 41 PM.jpg
IMG_1533.JPG

Contact is a crash detection helmet attachment for children. The device functions without a smartphone and integrates a crash detection system, GPS module, speaker, and app to notify caregivers in the event of an accident.

​

As part of a 15-person team in MIT's 2.009 Product Engineering Processes course, we created Contact. I led the crash detection algorithm development and testing, which included accurately classifying impact severity using industry standards such as the Head Injury Criterion (HIC) and Abbreviated Injury Score (AIS). I also presented our product at MIT’s 2.009 Final Presentation where I helped demonstrate Contact's functionality and potential impact.

 

MIT's 2.009 capstone course guides students through the product development cycle. Our project involved user experience design, hardware and software integration, and engineering validation. Contact is designed to improve safety for young cyclists by providing real-time crash detection.

Developing Contact allowed me to practice product design, engineering validation, teamwork, and presentation skills. 

​

One of the biggest technical takeaways was working on the crash detection algorithm. I implemented and tested impact classification methods using Head Injury Criterion (HIC) and Abbreviated Injury Score (AIS). I used Python to initially code our algorithm which was eventually implemented into our Rust program. Then through testing, I was able to test and process our sensor data to ensure detection accuracy while minimizing false positives. We created a drop rig to gather consistent crash data, had users bike around to assess the risk of false positives, and had user feedback from children on the form and placement of the device. 

​

Presenting Contact at MIT's Kresge Auditorium for our 2.009 Final Presentations was a valuable opportunity to practice technical communication. With our communications instructors, I played a large role in refining the scripts and the presentation to explain our technical engineering process to a broad audience. Presenting was an unforgettable experience.

​

Beyond these technical skills, I learned about teamwork and cross-collaboration. Working in such a large team meant communication was crucial between the components, and I found myself stepping up to h​elp out where needed to keep the project moving. ​With a fantastic team, I'm proud of what we created. 

Takeaways

  • ​Python

  • Verification testing

  • CAD

  • Technical communication

  • Teamwork

Process

Motivation
​

Biking provides children with independence, but it comes with major risks. Every year, hundreds of thousands of children suffer bike-related injuries. However, nearly half of children under 11 do not own smartphones, leaving children unable to call for help in these emergencies. While crash detection devices exist, they currently rely on smartphones to send emergency alerts. We saw this gap as an opportunity for innovation.​​​​

​​​​

​​

Goals
​​​

Contact was developed as an all-in-one crash detection system for young bikers that provides reliable safety alerts. We hope to give freedom for children to ride while still having access to emergency help when needed.

​

Design requirements based on background research and user interviews for the product included the following:

  • Accurate Crash Detection: Identify whether a crash has occurred while filtering out false positives

  • Real-Time Alerts: Give the location and severity of a crash to guardians

  • Independent of Smartphones: Allow children without smartphones to use the device

  • User Experience: Design a child-friendly interface with necessary communication functions

  • Helmet Attachment: Develop a lightweight, impact-resistant, and aesthetic mounting system

  • Weatherproof: Able to withstand dust and water (IP44 rating)

  • Convenient Charging: Last at least one week without charging 

Progress​

​​

Understanding Our Users

We conducted initial user interviews and surveys with parents to understand what they might need in a crash detection device. With this information, we could better define our functional requirements and determine what functionality was necessary and what would be a "nice-to-have". 

​

Hardware and Enclosure

After gathering feedback from parents, the hardware development focused on integrating the sensors we needed to satisfy our requirements. A microcontroller, IMU, GPS, and LTE module all needed to fit into a compact enclosure. This worked in parallel with other subteams to iterate and test different modules. 

​

The enclosure design had to protect the electronics, be a reasonable size to house the components, and maintain an aesthetic form factor. Many iterations of the form were tested for functionality and with children for user feedback on the looks and feel of the device. We also explored a variety of mounting strategies including adhesives, straps, and rigid brackets. This extensive testing led to the placement of the device on the back of the helmet and a Velcro strap solution to fit different helmet types. 

​

Crash Detection Algorithm

The crash detection system was one of the most critical components of Contact. I was responsible for designing and testing the algorithm. Through background research, we found industry-standard metrics that categorized crash severity, Head Injury Criterion (HIC) and Abbreviated Injury Score (AIS). 

​

The IMU sensor sampled acceleration data for us to constantly calculate and monitor the HIC score. If the score exceeded certain thresholds, it meant a crash was detected. ​These thresholds were defined by the relationship between HIC and AIS that has been studied in the automotive industry. This way, we could assign an AIS severity level – minor, moderate, or serious – based on calculated HIC scores.

​

In addition, we needed a head-sensing algorithm to differentiate a dropped helmet from a life-threatening crash. By using IMU data on whether the helmet was upright and slightly wobbling due to rider moments, we could ensure the helmet was being worn before confirming a crash. This head-sensing algorithm also helped keep our battery consumption low, as we could power on the device when the helmet was worn. 

​

App Integration

As the crash detection system started to be functional, we needed to create our app to communicate crash alerts with guardians. Using LTE for nationwide coverage, the Contact system updated the child's location every 30 seconds during normal use, and notified parents when a crash was detected. The child could also send simple notifications to the parent with a press of a button to say "I'm OK" after a crash was detected or "I have a non-crash emergency" if the child needed their parent at any time. We also enabled simplified two-way communication with preset messages to let their child know they have received the crash notification. ​After integration, we tested communication in different environments, which led to more iteration. 

​

Prototype Testing

Extensive testing was conducted to validate the system. We created a controlled drop rig for impact testing that helped us understand how the device responded to various crash scenarios. We dropped the device attached to a helmet and weighted head at a variety of heights to understand consistency in both the sensors and algorithm. We also used the drop rig to compare different potential placements of the device on the helmet to ensure that the device placement would not affect the sensitivity of the crash detection. 

 

We also tested our crash-detection algorithm against common false triggers and recorded the HIC scores of each action. For example, jumping, helmet drops, bike jumps, and more were performed to see if these could trigger an alert.

​

Through multiple iterations, we refined Contact's crash detection, communication, reliability, and ease of use.

​

Final Presentation

As we geared up for final presentations, I was elected as one of three presenters to highlight Contact's impact, technical design, and market to the audience. As a result, I played a large role in the slides and script iteration. We worked with our communications instructors to figure out how to best portray such technical information to the ~40k audience members that would be coming from a wide range of backgrounds. On December 9th, 2024, I proudly presented Contact at MIT's Kresge Auditorium in the 2.009 Final Presentations.​

​

bottom of page