T2D2 | AI Building Inspection
Cloud Inspection & AI-Driven Risk Detection for Infrastructure
Revolutionizing Maintenance with Geo-Tagged Imagery & Drones
Overview
T2D2's online inspection portal is a cloud software solution that allows building owners and their engineers to conveniently store inspection imagery all in one place. The unique dashboard including hundreds of images of geo-tagged damaged conditions and rank ordered by severity, gives you a detailed picture of where to focus preventive maintenance efforts to avoid higher downstream repair costs as well as potential safety issues.
Industry
Real Estate and Infrastructure
Country
United States
Project Summary
The original concept aimed to furnish engineers and building owners with a functionality enabling them to identify potential risks and damaged properties, empowering them to take prompt and informed actions. For image markings and shape detection, we utilized the Leaflet map library. Additionally, we incorporated custom features to address certain limitations within the library. We have done a complete unit test to make sure everything delivered to the client is thoroughly validated and meets the specified requirements. We have integrated numerous additional features, including the capability to display custom markers and images in the generated reports. Furthermore, we have implemented functionalities for managing drone records and facilitating the seamless upload of their associated images.
Case Study Research
Business Needs
- The conventional methods employed for image detection, storage, and assessment in traditional inspections are not only outdated but also prove to be inefficient.
- The requirement was to make such a platform through which engineers can have less time spent on field inspections and reduce the risk of safety for infrastructure.
- Adding AI to identify the potential risk of damage using images, videos and 3D models.
Challenges
- The main challenge was to handle the media uploading and managing them.
- Marking the detected damages to show the percentage of risks in damage on the image uploaded were somewhat hard to manage using the Leaflet library.
- When capturing any image of the building or house there was a need to get the dimension and the aspect ratio of them and show detected damages were quite hard.
Proposed Solution
- To Develop a system where one can view and manage detailed pictures of preventive maintenance efforts to avoid higher downstream repair costs as well as potential safety issues.
- To make the process smooth and reliable we have divided the features of the application into small yet efficient milestones.
