Robocist: Road Triage

Road Triage

Comprehensive, Objective Roadway Assessment
Better data for roughly half the cost!

Road Triage is a comprehensive, objective roadway assessment system that uses proprietary artificial intelligence to provide transportation agencies with curated pavement, lane line and traffic sign (coming soon) assessments for all the roads in their territory. Road Triage uses smart phones and fleet-based data collection to assemble road data, upload it to the cloud for analysis, and provide timely management reports. For roughly half the cost of current systems, Road Triage provides data that is current, comprehensive, and consistent. This holistic approach has the potential to provide significant annual cost savings along with access to geotagged, roadway assessment information that will help to improve roadway conditions, support more efficient maintenance, and support root-cause-analysis and predictive asset decay studies. A former NCDOT Transportation Secretary who initiated a Road Triage demonstration that ultimately exceeded expectation, stated that Road Triage has the potential to be a “Game Changer”.

Data Collection

Road Triage:
With Road Triage, data is captured by your fleet with windshield mounted smart phones. Fleet-sourced data collection utilizes the driven routes of a variety of vehicles ranging from standard cars to transit and school buses. The road data is collected, uploaded to the cloud, and analyzed. The artificial intelligence algorithm accurately identifies and assess pavement distress, lane lines and traffic signs (coming soon). Since the fleet vehicles are typically driven on the same route, results are provided on a monthly basis. Trouble spot areas with rapid degradation are identified and flagged. This allows users to quickly identify areas that need attention. Applying the monthly data, users can pinpoint when the roads degraded and identify the root cause. This supports more appropriate maintenance. Additionally, this same data can be used to predict which roads degrade faster and potentially when they will require repair.

Current Methods:
Typically, state and local DOTs identify primary and secondary roads to be assessed, then hire consultants/contactors to drive and walk selected roadways to gather data. There are two primary ways data is currently collected:
• Manually collected data requires hundreds of dedicated staff in each state to drive and walk the roads. This approach is dangerous for the workers walking unprotected on active roadways. Vehicles and unprotected staff have been struck, making it a hazardous job. Additionally, the process is expensive and time consuming, resulting in subjective, inconsistent roadway assessment data that may only be updated every few years.
• Most states also use LIDAR to capture data on a subset of the roads. A sampling of roads are selected for assessment to represent the larger unassessed road network. Although highly accurate, the fidelity is often lost when assessment averages are taken over stretches of road. LIDAR is also significantly more expensive than manual data collection and RoadTriage.

Data Quality

Road Triage:
The value of data is dependent on its quality. Road Triage’s proprietary artificial intelligence algorithms consistently and accurately identifies and assesses pavement issues, lane lines, and traffic signs (coming soon). When compared to the current manual data collection and assessment methodology, Road Triage produces better overall results by removing human subjectivity. Furthermore, the Road Triage AI models are regularly updated and retrained. A team of specialists combs through new data (or ‘new experiences’) to develop additional training material. The AI models learn from the new experiences. Our ‘continuous improvement cycle’ makes Road Triage even better with time. Road Triage provides actionable data which includes specific, color-coded, geo-tagged asset assessments. This level of detail provides better understanding of road conditions without having to visit in person. Data is viewable on Road Triage map tools and exportable for use in your preferred software.

Current Methods:
With such a large field data collection staff, data collection is often subjective and inconsistent. For example, some assessors may miss or deem road issues less severe than others. Once the data has been entered into the system, individuals then must review the data leaving another opportunity for subjective interpretations. The data often has gaps in coverage and is often outdated when used to develop maintenance plans.

Comprehensive

Road Triage: An advantage of Road Triage is that it can provide assessment data on all roads in a territory ranging from residential to highway at roughly half the cost of current methods. If state and/or local transportation organizations are responsible for maintaining all their roads in their territory, then it makes sense to have a holistic understanding of all the roads. Without it, there are gaps in understanding maintenance needs which may adversely impact planning and implementation. This can also lead to less equitable road maintenance and increased overall costs. Road Triage enables users to assess all their roads for no additional cost. Users can even include the most current road data from multiple months for cases where areas are driven less frequently.

Current Methods:
At best, data are collected on a rolling basis to provide annual data on most primary and secondary roads. Considering the magnitude of the road system, it is not economically or logistically feasible to capture data on all roads. This results in using outdated information with data-gaps to develop maintenance plans. Additionally, even in optimal scenarios that include annual road data, it is not granular enough to conduct root cause analysis or predictive modeling. Both of which would support a more accurate, more efficient short and long-term maintenance planning. There would be less surprises and a better understanding why roads had issues. This would in turn support more informed approaches to repair and maintenance.

Lower Cost

Road Triage:
Better data at roughly half the cost. High-quality road data at a low price is the cornerstone of our vision. The current process requires expensive LIDAR equipment and/or a large dedicated staff in each state to drive and walk the roads. This approach is dangerous for the workers walking unprotected on active roadways. Vehicles and unprotected staff have been struck, making it a hazardous job. Worker attrition is also an issue creating pressures for ongoing recruitment and training. Using Road Triage, road data is collected while fleets drive their typical routes. Hence there is no special, additional effort required for the majority of the roads. And for those areas not normally covered by fleet operations, the labor effort required is merely to drive the roadway. There is no need to exit the vehicle and manually evaluate the roads. The reduced data collection and processing effort translates into a lower cost with increased data quality and frequency. Additionally, the safer work environment may help to reduce associated insurance costs.

Curent Methods:
The current process is extremely labor intensive, requires dedicated vehicles, has health and safety issues, and does not cover the entire territory. And since pricing is typically based on roadway miles, DOTs are forced to focus on the roads they feel will have the biggest impact as opposed to the overall roadway system that includes feeder and residential roads. Considering the current economic scenario, effective cost-saving approaches are needed for data collection as well data applications that can have cross-departmental impacts.