Hackathon at The Southern Methodist University.

Inspiration : The inspiration behind this project is to revolutionize commercial building management by harnessing AI and machine learning to provide real-time insights, predictive maintenance, and user-friendly interfaces for building managers. We aim to address the challenges of asset management in the commercial buildings industry and create a more efficient, cost-effective, and sustainable approach that benefits all stakeholders.

What it does : Our project is a comprehensive asset management solution designed for commercial buildings. It employs advanced machine learning algorithms to analyze data from various sources, including operational logs, maintenance history, and seasonal parameters. The system provides real-time status monitoring of critical assets such as HVAC systems, elevators, and electrical panels while also predicting maintenance needs and potential failures. Its user-friendly interface facilitates data-driven information insights for building admins, ensuring efficient resource allocation and proactive maintenance. The project aims to enhance operational efficiency and reduce costs while optimizing the management of commercial building assets.

How we built it : We collected data from various sources including the CBRE challenge provided data. We processed these different sets of data and ran some advanced ML algorithms to get insights on the data which was further utilized for anomaly detection and prescriptive analytics. We computed parameters like Avg. Service time, and leeway period for different machines on the basis of various factors. In the end, we built a platform that consisted of several dashboards to provide continuous monitoring of the assets that a user is in charge of and also raised alarms and notifications via emails when any particular appliance requires attention.

Challenges we ran into : We were lacking on suitable data to train the models and get insights from them. We had several UI and backend integration issues. We had challenges in selecting the best model for some prediction use cases.

We learned a lot about team work ,project management and technology expansion. Got to learn new techniques to use in the data insights speculations.

What's next for Lens for CBRE A full fledged one stop solution for all the building administration related work like work order raise, vendor switch, machine health monitoring, service recommendations, etc. For starter it would require a much larger amount of data set to have a better reach and accuracy on the real world scenarios.