Scout Robotics' rail sensors measure track within 1 mm at up to 74.5 mph
Scout Robotics has launched Scout Series X and Scout Series T, infrastructure inspection systems that mount on nearly any rail vehicle and collect data during ordinary runs. The systems can operate up to 74.5 mph with measurement accuracy of ±1 mm. Scout’s pitch is that repeat passes over the same assets enable change detection and support a “state of the network” report aimed at capital planning.
This story was produced through MarketScale. See how Transportation teams put it to work with Partner & Channel Enablement.
Key takeaways
Scout Series X and Series T mount on nearly any rail vehicle and measure track to plus or minus 1 mm at up to 74.5 mph, collecting inspection data during regular service.
For a sensor that rides revenue trains, traffic becomes the inspection schedule, so deciding which vehicles get equipped on which routes is part of the inspection plan.
The output is a capital-planning report with no specific repair recommendations, so the railroad has to define the assets, thresholds and segments it wants measured before the first sensor goes on a car.
Free workspace
Turn your Transportation expertise into content.
Record interviews, organize footage, and write with AI on a free trial of the MarketScale platform for qualifying companies. No demo required, no credit card.
Scout Robotics wants the freight car that's already on the schedule to handle track inspection too. The Philadelphia company has introduced Scout Series X and Scout Series T, two infrastructure inspection systems that mount on nearly any rail vehicle, whether rolling stock or hi-rail, and record conditions during ordinary runs. Trains, which reported the launch, said the equipment points down at track and right-of-way assets and forward to capture lidar and imagery of the terrain along the line.
The headline specs are speed and precision. Chief executive Tejas Agarwal told the publication the systems work at up to 74.5 mph (120 km/h) with measurement accuracy of plus or minus 1/32 inch (1 mm). His pitch is mostly about where the sensors ride: the systems, he said, turn "vehicles already moving across the network into infrastructure intelligence platforms."
For a maintenance-of-way chief or a transit agency's asset manager, the harder question is what a steady stream of passes from revenue equipment adds to the inspection program already in place, and what the railroad has to decide before that data becomes a capital plan.
Scout is young, though its founders are not new to the sector. Agarwal and chief technology officer Arjun Nanda, both University of Pennsylvania graduates with backgrounds in autonomous vehicles and geospatial AI, have been involved with transportation startups over the past 10 years. They founded this one in 2023 as Sahay AI. Two years later, in 2025, the business took the Scout Robotics name and relocated its headquarters to Philadelphia. That same year it introduced SPARK, a lightweight system that runs on batteries and inspects both rail and wayside assets. Series X and Series T build on that line.
Why the second pass matters most
Trains framed the market plainly: suppliers mostly address track inspection, some Class I railroads have built their own hardware and software, and the systems in use vary in how well they follow the condition of individual assets at precise locations over time. That last gap is where Scout is aiming.
The company builds digital twins and fuses data from several sensors, with modes including visual, RGB image, vibration and acoustic, so users can hunt for specific defect types. The capabilities it lists include visual defect detection, track geometry, thermal anomaly detection and predictive asset modeling, along with mapping of vegetation, tunnels and clearances. Users see results in real time or on a dashboard, and Scout also sells wayside devices that inspect trains as they pass. The catch, which Trains noted directly, is that making sense of all this depends on sophisticated AI models and comparison across multiple runs.
Change detection is the core method here. A single run produces a snapshot; Scout's assessments rely on comparing the same assets across repeated runs to see what moved, wore or grew since the last pass.
That dependency explains the mounting strategy. A sensor on equipment that covers the same segment day after day builds a condition history for each tie or component, which is exactly what change detection feeds on. Put another way, the accuracy spec earns its keep on the tenth pass, when a 1 mm shift is a trend rather than noise.
Field record: 250 miles in 3D
Agarwal gave Trains a running tally of what Scout's systems have done in the field so far.
Scout Robotics field totals to date
Trains
Those totals span freight, heavy rail, light rail, commuter rail and yard settings, and include more than 250 miles of infrastructure reconstructed digitally in 3D. Five operating environments is a broad proving set for a company founded in 2023. The 250 miles is the number to hold onto, because it describes a footprint measured in segments and is the right yardstick for asking whether the models have seen track like a buyer's own.
Dividing two of the company's totals ourselves, 500 terabytes over 1,000 hours, gives about half a terabyte per field hour. That is our back-of-envelope average across products and settings, not a Scout specification. For a railroad CIO, it is still a reason to settle early where raw data lives, who keeps it and for how long, since change detection only works if past runs remain available for comparison.
The case for Scout is weakest where a railroad already has it covered. Class I carriers that built inspection systems in-house set a high bar for any outside vendor, and adding another data stream carries its own integration cost. The opening looks wider for operators without that engineering bench, which could include smaller freight carriers and transit agencies.
A network report for the capital budget
Nanda told Trains the systems are best suited to capital planning. Scout does not recommend specific actions; it delivers a "state of the network" report built around whatever the railroad wants to examine. After capture, data runs through the company's proprietary AI models and is presented on the Scout operating system, formatted to customer requirements, compliance requirements and regulatory standards.
Agarwal's list of uses runs wider. He said the data is delivered through the Scout platform "to support earlier detection, maintenance prioritization, reliability, and capital planning." Read together, the two executives suggest capital planning is the lead use, with the same condition history meant to serve maintenance teams working on shorter horizons.
That split keeps the engineering judgment with the railroad. It also puts the scoping work on the buyer. The report reflects the questions asked, so the asset classes, thresholds and segments that matter need to be written down before the first sensor goes on a car.
On Nov. 17, Railway Age reported that the North Carolina Department of Transportation is putting $13.3 million into freight rail infrastructure upgrades. The work covers the Port of Wilmington, three economic development projects and 12 short line railroads. On Nov. 16 it reported that Kentucky Gov. Andy Beshear announced $7.3 million in grants for six railroad infrastructure projects. Operators competing for that kind of money, or spending it, could use a condition report that ranks segments by measured change as supporting evidence, as long as the railroad trusts the comparison behind it.
Much of that money points at short lines. Railway Age headlined the Kentucky grants as short line infrastructure awards, and on Nov. 16 the American Short Line and Regional Railroad Association reported the 51st co-sponsor of S. 1532, the Railroad Track Maintenance Credit Modernization bill. Short lines sit squarely in the group that could lack an in-house inspection bench, which would make them a practical test of whether a vehicle-mounted sensor package and a network report can fill that role.
Transit capital is moving as well. Railway Age reported on Nov. 13 that NJ Transit's board authorized $65.6 million for capital improvements to the Hudson-Bergen Light Rail system, the first allocation of a planned $250 million investment. Light rail is one of the five environments in Scout's field record, and an agency sequencing a multiyear program in tranches is exactly the buyer a repeat-pass condition history is pitched at.
Ask Scout for two consecutive runs over the same segment and watch how the platform flags change on a single tie or component. Then ask how those flags map onto the railroad's own compliance format, since the company says it shapes output to customer and regulatory requirements.
Busy tracks mean more passes
Traffic is running high. Railway Age's November issue carried the Association of American Railroads' read: U.S. railroads originated nearly 295,000 intermodal containers and trailers per week in September, which made the third quarter the highest-volume quarter ever for U.S. intermodal. Total carloads have grown year over year for nine straight months, and in September, 17 of the 20 carload categories the AAR monitors had year-over-year gains, the most in five years.
The AAR added that, excluding coal, carloads are growing twice as fast as the overall economy, and that the growth runs across the industrial spectrum rather than a handful of isolated sectors. Broad-based freight could mean extra trains on more of the network, not only on a few intermodal corridors.
For a system that rides revenue trains, traffic volume becomes the inspection schedule: every additional equipped train over a segment is another pass to compare. That holds only if the railroad equips enough vehicles on the right routes. Fleet assignment, normally a transportation question, becomes part of the inspection plan.
Appetite for these tools has been building for a while. In Logistics Management's November 2025 issue, Michael Levans recounted the magazine's annual rail and intermodal roundtable, where group news editor Jeff Berman summed up panelists' view: railroads trail trucking, air and ocean in automation but are investing in tracking, predictive analytics and customer-facing tools, with AI adoption still early and promising.
The same panel said railroads are beginning to use predictive analytics and AI to optimize network fluidity and better allocate assets, and that rail performance metrics are harder to obtain and less often analyzed publicly. A per-segment condition history is one kind of metric that sensors riding revenue trains could make routine, at least inside the railroad's own planning office.
The advisory bench and the next signal
Scout also added two advisors. Matthew Igoe is president and chief operating officer of R.J. Corman and a former BNSF chief operating officer. Don Itzkoff founded Transport Strategies and was deputy administrator at the Federal Railroad Administration. They join an existing advisory group that includes Ray Betler, who chairs L.B. Foster and previously served as Wabtec's CEO; Leslie Richards, formerly Pennsylvania's secretary of transportation; Randell Iwasaki, a past director of the California Department of Transportation; University of Pennsylvania professor Rahul Mangharam; and Sondera chief executive Josh Devon, whose company focuses on security for AI agents.
Read as a roster, the bench maps onto the buyers Scout is courting: freight operations at Class I and short line scale, federal safety oversight, state transportation departments and a major track-products supplier. That suggests the company expects its sales path to run through public agencies as well as private railroads.
The next marker is a named fleet. A freight railroad or transit agency announcing Series X or Series T on its own vehicles, and saying which decisions the network report feeds, would show whether riding along with daily traffic can earn a line in a capital plan.
Sources
Featured companies
Your experts belong here
Every story in MarketScale Transportation starts with a company putting its fleet managers, logistics engineers, and safety leads on the record. Buyers are already reading this topic. The only question is whose experts they find.
Fleet and logistics buyers compare quietly, and your operators become the evidence that settles it.
About the author
The MarketScale Newsroom reports on the companies, technologies, and trends shaping 16 B2B industries. It turns primary sources and expert commentary into clear, useful coverage for the people doing the work.