Mapillary Imagery of Timmons Group Capturing in Clear Creek CO.
Transportation agencies face a near-impossible challenge: maintaining hundreds or thousands of miles of roadway with limited resources and staff. Over time road surface conditions degrade due to general usage and weather conditions. The agency’s staff can't be everywhere at once, making it incredibly difficult to identify cracks, potholes and degrading roads to keep up with maintenance. Establishing general awareness of pavement conditions and distresses alone can be a major undertaking. Many agencies even lack an authoritative inventory of assets, such as signs, guardrails, gutters, and more, further complicating related maintenance tasks.
But what if they could be provided with a snapshot in time for their roadway network one that doesn’t just show data on a map or spreadsheet, but also offers a virtual, explorable copy of the roads?
Timmons Group's workflow begins with a custom-equipped truck that serves as a mobile data collection platform, guided by digital field maps containing GIS representations of assigned roadway segmentation and outfitted with an array of SSI laser sensors, as well as Mosaic 51 360-degree camera system. The team captures both high-resolution georeferenced imagery and pavement roughness data through a predefined route, ensuring maximum capture of the state roadways.
Timmons Group’s GIS staff process the panoramic imagery and then upload to Mapillary using the Desktop Uploader. Staff and clients can then explore the latest view of the applicable road system through the Mapillary web app and Esri integrations like ArcGIS Pro and the online Experience Builder (such as the Oriented Imagery Widget using Mapillary’s catalog).
Timmons’ GIS team prepares web maps and apps to be utilized by our internal data collection team. These apps enable technicians to digitise and to systematically review the road network, capturing pavement distresses and cataloging assets using Mapillary imagery. In some cases, manual asset extraction is supplemented (or substituted) by a faster approach in which a Python script is utilized to pull Mapillary’s computer vision derived assets from the site’s API. While assets of interest vary from client to client, traffic signs, traffic lights, utility poles, and manholes are commonly provided.
Typically all Sign and Object types are extracted from within a pre-defined bounding box as a json file. The json is then converted to a feature class, clipped to the project boundary, symbolized to match Mapillary’s symbology, then published to a feature service. This service is typically used in conjunction with other project GIS data to present a comprehensive, aerial view of the roadway network, down to the individual stop sign and manhole.
Several example images of the Mapillary API requested object/sign outputs within custom web maps. For ease of use, Signs are typically separated from Objects into sublayers within the feature service. Visibility of these layers is typically set to only appear when zoomed-in to about street-level, to avoid visual clutter.
Timmons Group then builds a series of web applications which combine editable GIS maps with a custom Mapillary imagery viewer panel. These applications reference the Mapillary API to enable filtering of images by area, Timmons Group’s Mapillary username, and/or date range. Images are then presented on a mini-map that users can utilize to jump between areas or play through a sequence of images. Providing these images alongside the editable web map, allows users to visually assess the pavement and collect individual distress points with attribution, such as the type of distress,the severity, and the extent.
Example of a custom web application developed by Timmons Group for Pavement Distress Collection. On the left side is the Mapillary imagery viewer with a mini-map presenting all filtered project images. The right side is the editable web map which can be synced to the Mapillary image location, to efficiently drop distress points and collect attributes
Equipment from SSI enables sensor-derived roughness measurements (IRI) from the pavement truck and pavement distress points collected using the custom GIS/Mapillary web app are then fed into specialized software, Paver, to determine Pavement Condition Index (PCI) based on industry standards through a series of complex calculations. This metric informs the client on the overall health of each assessed road segment.
Once virtual data collection is complete, the GIS team is able to prepare final deliverables that bring everything together in ArcGIS Online. An assortment of web maps, ArcGIS Experience Builder applications,dashboards,and Mapillary’s imagery enable clients to access and interact with pavement conditions, distress locations, network-wide statistics.
Example of several typical configured web maps and dashboards developed by Timmons Group for pavement data delivery. Dashboards show clear metrics on PCI, RQI, asset counts, and more. Filters enable users to interactively sift through data. Typically, a series of maps are prepared, though the ‘All-In-One’ map contains all of the data collected for the project.
One notable product provided to clients is the Pavement Viewer Experience. This ArcGIS Online Experience Builder tool consists of several pages. The first page combines GIS data, such as road segments symbolized by color to show PCI score categorization (from ‘Serious’ to ‘Very Good’) alongside assets and distresses with a non-Mapillary pavement imagery; this imagery is not 360 and is primarily focused on showing the assessed pavement. The second page of this Experience utilizes interactive Mapillary 360 imagery via the Oriented Imagery Widget.
Available imagery locations are shown above the assessed segmentation, allowing clients to easily identify relevant areas for viewing - to enable a filtered view of the Mapillary imagery, this page links to a custom webpage that pulls in Mapillary images via API request.
(Above) Example of the Experience Builder application, showcasing GIS data in conjunction with Mapillary’s imagery catalog via Esri’s oriented imagery widget. Check out the EXB here. Please note that for sharing purposes, data within this app has been limited – typically there are more layers and widgets to assist in data review.
By pairing Timmons Group's innovative pavement assessment methodology with Mapillary's integration with Esri, a snapshot of clients' network of owned/operated roadways can be achieved. The combination of GIS data and imagery creates a virtual window from which users can literally view their pavement in conjunction with its conditions. This knowledge allows clients to allocate funding for maintenance operations, establish catalogs of assets, and run future analyses.
The true power of this approach lies in its ability to transform abstract data into understanding. Usually, a PCI score of 65 is just a number on a spreadsheet but combining that score with imagery and supplemental GIS data contextualizes those scores. Users can see the actual cracks, deterioration, and findings within the captured imagery as well as view all the collected data within the local geography on a map.
For agencies managing rural terrain and remote roadways, this visual validation is invaluable. Officials can examine pavement conditions on rural routes without dispatching crews for time-consuming site visits. And when maintenance priorities and funding are debated, decision-makers can take a look through the virtual window to review road conditions together.
"Seeing is believing" extends beyond internal decision-making. When presenting infrastructure needs to county commissioners, town councils, or the public, the ability to show, not just tell, can make all the difference. Visual documentation lends credibility to funding requests and helps non-technical stakeholders understand why certain roads require immediate attention while others can wait.
As it was referred to earlier, both the Mapillary imagery and pavement data captured today combine into a temporal snapshot. What was current at the time of collection becomes a historical record for later, enabling agencies to track deterioration over time, document pre- and post-construction conditions, or even revisit locations months later to investigate reported issues without returning to the field.
Ultimately, Timmons Group's utilization of Mapillary in conjunction with GIS technology empowers public and private clients alike to better see and understand their roadways. The tools and data provided help to keep roads safe and pleasant to use, whether through the beautiful mountains of Colorado or the historic cities of Virginia.
Timmons Group has provided pavement clients across the US, ranging from 100-1,500 mile areas, with a variety of GIS-based technology solutions and data. Check out the Pavement Portal to learn more about our approach and to see the growing project catalog.
Check out what else Timmons Group's technology division can do for you here!
Andrew is a Technical Analyst on Timmons Group’s Geospatial Services team, providing expertise in GIS solution design, data analysis, and system configuration. He has over seven years of experience supporting public- and private-sector clients, with project work spanning transportation planning, asset management, field data collection, and analyses. His experience includes modernizing field assessment and asset inventory workflows, performing analyses on large and complex datasets, developing customizable reporting solutions, and supporting emergency response GIS efforts. Andrew is also involved in strategic planning, needs assessments, and data strategy initiatives, where he contributes technology recommendations, KPI development, and guidance on data quality, security, and governance. He is well versed in business analyses, desktop/online/mobile GIS solution configuration, and the ability to communicate with and train stakeholders on complex GIS topics.
Brian is a GIS Solutions Architect and Project Manager on Timmons Group’s Geospatial Services team providing expertise in GIS. He has worked with numerous clients in both the public and private sectors. His project experience ranges from performing mission-critical GIS tasks while serving in the military here and abroad, to assisting DOTs in solving business problems, as well as helping local governments improve workflows. He is well versed in implementing both desktop and web GIS workflows using ArcGIS Pro, ArcGIS Online, and ArcGIS Enterprise for organizations of all sizes.