Data & Tools
- The dataset is a 3D LiDAR point cloud collected from a vehicle driving through the site.
- Each point represents a measured location in 3D space and may contain information such as:
- X, Y, Z coordinates
- Laser intensity
- Return information
- RGB/color information
- Other LiDAR metadata
- Files (LAS 1.4, point data record format 7) Run 1 Laser Left - Run 1 Laser Right Run 2 Laser Left - Run 2 Laser Right
- Site: Mannford Oklahoma
- System: Trimble MX9 - two scanner heads
- Captured by: WSB
Policies and procedures are just as important to infrastructure owners as using innovation and technology. A key rule of engagement for this challenge is to document a repeatable policy/procedure for an infrastructure owner to follow in developing your model.
Added challenge would be to document a policy/procedure for improving on your model as well as replicating.
Students are not limited to the Business Center AI/ML model for this challenge, if they find other AI processing helpful/complimentary to their work on the challenge they are open to leverage other tools to assist.
Judging
- Scoring Metrics (100 Points Total)
- Extraction Accuracy (30 pts): Precision in geospatial extraction accuracy.
- Policy and Procedure (30 pts): Documented method for repeatibility for the assets.
- Attribution Completeness (30 pts): Consistency of classifcation of the assets.
- Innovation: The Enhancement Challenge (10 pts): Documented method for enhancing model approach for assets.
Submission Deadline
All submissions must be finalized on Devpost by Sept 11, 11:00 AM, before judging begins.
Code of Conduct
All participants, mentors, judges, and organizers are expected to help create a safe, inclusive, and harassment-free experience for everyone, regardless of background, identity, or experience level. Harassment, discriminatory jokes, disruptive behavior, or any form of intimidation will not be tolerated and may result in disqualification or removal from the event.