Atlas

Inspiration

Safety. Unfortunately, there are a lot of school shootings and robberies occurring all over our nation. We want to take a step further in preventing these unfortunate events from taking place in the future, saving as many lives as possible.

What it does

Analyzes video footage to understand any suspicious behavior and immediately notify the appropriate parties as well as take safety and preventive measures.

How we built it

We used YOLO for live motion tracking and object detection. We used Nemotron models to interpret behavior patterns and lookout for suspicious or unsafe situations. We used OpenClaw to evaluate the best next steps for preventive measures in these situations and we used NemoClaw to ensure the decisions we make as preventive measures are appropriate to the situation, essentially further risk assessment.

Challenges we ran into

Securing footage for suspicious situations, understanding behavior through computer vision, and implementing guardrails for risk assessment, as well as accurate object detection (mainly due to ambiguity in suspicious classification).

Accomplishments that we're proud of

We think this product can really help save a lot of lives by predicting dangerous circumstances beforehand and acting accordingly. The tiniest amount of time in leads of a dangerous situation can help save a life, and we hope to be a part of the solution!

What we learned

AI gave us a lot of resources to build with, to develop solutions to existing problems. We chose what was most meaningful to us and the amount of work we were able to accomplish in a mere 24 hours astonished us and taught us how valuable our contributions can be in the future!