We protect bridge users from vessel collision using real-time transponder (AIS) tracking and computer vision.
We combine vessel trajectory forecasting with AASHTO bridge impact analysis to continuously assign a threat level to each vessel within 30 nautical miles of the site.
When the containership Dali collided with the Francis Scott Key Bridge, police had only two minutes warning to close the roadway. Six people died and the collapse cost society an estimated $5 Billion. The Port of Baltimore was closed for months and 8,000 workers lost their jobs.
Photo: The containership Dali with portions of the collapsed Francis Scott Key Bridge across its forward deck on March 28, 2024. (Source: NTSB)
After an intensive investigation, the National Transportation Safety Board (NTSB) identified 68 bridges across the U.S. that are potentially vulnerable to vessel collision risk.
On 18 November 2025, NTSB recommended to the owners of these bridges:
"Recommendation 15. As part of your short-term bridge risk reduction and mitigation strategies to protect the traveling public, evaluate the need for and, if appropriate, incorporate motorist warning systems capable of activating when a threat is identified and immediately warn and stop motorists from entering the bridge." (Reference: 4 hr 11 min 35 seconds)
Roebling Labs protects bridge users from vessel collision using real-time transponder (AIS) tracking and computer vision. We combine vessel trajectory forecasting with AASHTO bridge impact analysis to continuously assign a threat level to each vessel within 30 nautical miles of the site.
Roebling Labs has developed a prototype Vessel Collision Warning System available here: vesselcollision.com
The vessel data for the prototype is pulled from AISStream.io and auto-refreshed every 30 minutes for demonstration purposes. Production systems deploy a local AIS receiver at the bridge site for continuous tracking. Enhanced systems use computer vision for redundancy, protection against attack, and incident logging. In the case of false alarms, feedback from the computer vision incident logs facilitate reinforcement learning to update the threat assessment algorithm.
Figure 1: Screenshot of an example vessel collision warning system by Roebling Labs
Figure 2: Risk Matrix for Assessing Vessel Threat
Contact scott@roeblinglabs.com to demonstrate the vessel collision warning system for your bridge.
See the User Manual for detailed screenshots and explanations for how your vessel collision warning system (vesselcollision.com) will protect the traveling public at your bridge site.
Roebling Labs protects bridge users from vessel collision using real-time transponder (AIS) tracking and computer vision. We combine vessel trajectory forecasting with AASHTO bridge impact analysis to continuously assign a threat level to each vessel within 30 nautical miles of the site.
Roebling Labs was founded by Scott Snelling, P.E. in 2024 with startup funding from Bentley Systems, the maker of Microstation CADD software. We also received an early grant from the BUild Lab startup ecosystem at Boston University. Our founding vision was to build innovative bridge assessment tools.
Roebling Labs is now a resident company at The Engine. The Engine is a non-profit startup incubator that was "Built by MIT to accelerate early-stage Tough Tech companies." In this context, Tough Tech is defined as "transformational technology that solves the world's most important challenges through a convergence of breakthrough science, engineering, and entrepreneurship."
Roebling Labs offices are located within The Engine at 750 Main Street in Cambridge, Massachusetts where we are co-located with other Tough Tech startups and venture capital firms.
In response to NTSB recommendations resulting from the Francis Scott Key Bridge collapse, Roebling Labs is fully committed to leading the bridge industry in providing vessel collision warning systems to protect the travelling public.
Scott Snelling, PE*, MBA is the CEO of Roebling Labs LLC. He has 20+ years of bridge engineering experience having led complex bridge projects across the U.S. and abroad. Scott co-authored portions of the Bridge Engineering Handbook. He led projects to recommend revisions to AASHTO bridge standards, as well as AASHTO bridge maintenance, evaluation, and inspection standards. With Roebling Labs, Scott has developed and tested numerous emerging bridge assessment technologies.
Prior to founding Roebling Labs, Scott began his career at WSP, U.S. Army Corps, and Hardesty & Hanover. He has degrees from Columbia University in the City of New York (M.S. Structural Eng.), Rose-Hulman Institute of Technology (B.S. Mechanical Eng.), and Boston University (M.B.A.). *(Professional Engineer WA #43339)
Shaun Meredith is a computer vision expert and former Navy submarine officer. He employed computer vision as a plasma diagnostic within MIT's versatile toroidal facility group of the Fusion Center. He earned his M.S. from MIT. Now, as the Co-Founder and Chief Technical Officer of Omnic.ai he applies computer vision and artificial intelligence to coach e-sports athletes. In his free time, Shaun embraces the marine environment by racing sailboats.
Email scott@roeblinglabs.com to protect your bridge users today.
Pilot program available - deploy at your site with no upfront commitment. Pay only after successful validation.
Get updates on vessel collision warning system deployments and best practices for using computer vision on bridges.