We are a bunch of friends who really miss going climbing at our local gym. Since the beginning of the pandemic, we only went on a few occasion, because it was closed so many times. One of the team member built himself a climbing wall at his girlfriend's house, so he could do what he loved when gym where closed! However, building a setup at home can not replace the experience of climbing at the local gym with your friends, as such, we add to find a way to keep the gyms open!
We realized that we already have all the tools we need to climb in a sanitary environment, the only problem was at the time of the execution! Even on their best behavior, people did not fully respected the COVID rules and it is impossible for the staff members to warn them all. They would need to be everywhere and see everything. But what if it was?
What it does
CHAMO.io is a solution to monitor the rooms the trafic in your buildings, keeping safe your client and your company.
With the video feed of your security cameras, we perform object recognition to detect people in each zone of your buildings, allowing you be confident you always respect the COVID health mesures. The application has many monitoring features, such as setting alarms when a zone has too many people, or understanding through insightful schemas where you are the more at risk of being infected, based on trafic and last cleanup time.
With that much knowledge about the most visited zones, you will be able to optimize your employee schedule by making them work where it matter, and please your client by providing them better services.
How we built it
The application uses state of the art cloud services, thanks to Amazon Web Services.
Our application has a front end and a back end and is hosted on AWS.
The back end of our application is made of two parts. We have an API storing live and historical data about your building's traffic. It is entirely managed by Amplify, a IaC service. The whole definition of our API is described in code, which makes it easy to control our architecture. The streaming services for the video cameras uses AWS IVS service, the technology behind the successful streaming application, Twitch. As such, video are easily accessible and have a low latency.
The front end is comprised of differents views to monitor your building and manage all your schedule's restrictions. In the monitoring view, you have acces to a live feed of all your building's traffic with human detection enabled by a YOLO model. Also, this view provide some simple visualizations about how many people there is in each room for some specific periods in the past days.
Challenges we ran into
Funny enough, even if our teams members had a few tricks up their sleeves, we encountered a lot of obstacles when it came to do front end development, which is often time one of the first skills you learn in software engineering. With the help of our front end sempai, we managed to overcome the difficulties and learned a great deal along the way.
Accomplishments that we're proud of
- Manage to implement an in-browser object recognition
- Being one of the first users of a new AWS service (IVS - 7 months old)
- Using simple, yet powerful visualization tools developed specifically for React
What we learned
Front end development is an overlooked skill in startup environments. Its has a strong creating power that allows you to design rich applications. It's ever newer technologies makes it an interesting domain in computer science.
Hackathons are more than a competition. It is a time to have fun and share our love for software engineering.
What's next for CHAMO
- Building the alert service with AWS SNS service
- Moving the object recognition software on the cloud
Notice: web page takes time to load livestream