Facilitrack

Augmented Reality Asset Analyzer
Screenshot of Main Dashboard
Screenshot of Search Page

Inspiration

We were inspired by the CBRE challenge to build an app that not only leveraged machine learning and analysis to manage commercial real estate assets, but to include unique features that greatly enhance the end-users experience.

Our augmented reality feature was inspired by Google Lens and ARUtility as well as our wish to have an app that could help direct us to where confusing machines are located.

What it does

Facilitrack tracks building assets (HVAC units, fire alarms, elevators, etc.) in a commercial facility. It allows building owners and tenets to track the upkeep of these machines and locate where they are located using augmented reality. Facilitrack also gives repair recommendations, anomaly alerts, and visual overviews of the entire system.

How we built it

The front-end is made with Sveltekit and Streamlit. The backend consists of Sveltekit API endpoints, a MongoDB Atlas deployment, and several machine learning applications running in Python. Google Cloud and Twilio are used for hosting and the notification system.

Challenges we ran into

Due to the variety of technologies used in this project, integrating them was complicated. Most parts communicate using JSON REST APIs, but some communicate through iframes. We managed to get everything to work well together, but it took a significant amount of time.

Accomplishments that we're proud of

We are proud that we accomplished what we set out to do. We implemented nearly every feature that we planned in the beginning and even had some time to add more. Overall, we are proud of our project.

What we learned

Everyone on the team learned something new. Many technologies in the stack were unfamiliar to some team members such as Sveltekit and Streamlit. However, by working together, we were able to work together to quickly teach each other everything we needed to know. Now, we are all comfortable with the technologies that we each used.

What's next for Facilitrack

The next logical step is cleaning up, refactoring, and making Facilitrack more stable. This is a 24-hour hackathon project, so the next step is fixing all of the problems we know exist but didn't have time to fix.