Untitled
The Class I railroads are not being honest about Automated Track Inspection (ATI): The railroads do not need waivers or suspensions of federal safety regulations in order to test or use the track inspection machines. The waivers and suspensions of regulations sought by the railroads concern the frequency of human track inspections. These human inspections still need to be required to ensure track safety because ATI does not pick up on the vast majority of track defects. Only 26% of all track related defects can be detected by ATI, the other 74% can ONLY be detected by a human being.
The Class I railroads are actively trying to reduce required visual track inspections, a move that would result in more track defects undetected, more derailments, and a less safe railroad system nationwide.
What Brotherhood of Maintenance of Way Employes Division (BMWED) Supports: BMWED believes that the rail industry should voluntarily or be required to adopt a higher ATI inspection frequency than currently required to catch more track geometry defects, while maintaining human visual frequency of inspection requirements to catch other track defects. BMWED also believes FRA should clearly state that increased ATI frequencies are to supplement, not replace, the required human visual inspection frequencies.
Current Status of ATI
In December 2025, FRA approved an industry-wide waiver allowing freight railroads to reduce federally required visual track inspections under 49 CFR § 213.233, contingent on the use of Track Geometry Measurement Systems (TGMS)-also known as Automated Track Inspection (ATI). The waiver allows a reduction in visual track inspection frequency from twice weekly to once weekly. FRA indicated that it would evaluate performance after the first year to determine whether any further reductions in inspection frequency may be appropriate. The Federal Railroad Administration’s ATI waiver permits railroads to reduce federally required visual track inspections only if they first satisfy specific waiver conditions intended to ensure that automated inspection can be evaluated for safety equivalency. This waiver represents a fundamental negative shift in rail safety policy by permitting automation to replace certain human inspection functions. (BMWED-IBT President Tony Cardwell Statement on Automated Track Inspection (ATI) Waiver Decision)
74% of defects identified by human track inspectors involve conditions that ATI systems are not designed to detect, including component-level issues such as fasteners, ties, ballast conditions, and other structural deficiencies such as track washouts or obstructions. This fact reinforces that ATI is not a comprehensive inspection method and cannot serve as a direct substitute for trained personnel.
Two Class I railroads have now submitted notifications of intent to operate under this waiver. A review of those submissions shows that the central problem is not simply disagreement over policy, but that the data provided by both railroads is too incomplete, too fragmented, and too difficult to independently verify to support confidence in reduced human inspections. (regulations.gov)
Union Pacific Railroad (UP) proposes ATI operations over approximately 2,198 track miles and 1,363.74 route miles, divided into 10 groupings.
BNSF Railway proposes ATI implementation across approximately 23,932.75 track miles and 16,682.28 route miles, divided into broad “North” and “South” reporting territories.
Union Pacific’s waiver went into effect on March 29th, while BNSF’s waiver on their South territory went into effect on April 5th. BNSF’s waiver on their North territory will go into effect on July 5th. For the portions of UP and BNSF’s territory where the waiver has gone into effect, we have received confirmation that they have reduced visual track inspections to once a week. (BMWED General Chairpersons Request Clarity on BNSF’s ATI Waiver Plans)
These Inspection reductions increase the risk of derailments in our communities. Inspection reductions are now being implemented before the railroads have provided a complete and independently verifiable data record demonstrating that the waiver conditions have been satisfied.
Interactive ATI Territory Map
Compliance With the Waiver
A review of the railroads’ submissions raises several significant concerns regarding whether the waiver conditions have been satisfied.
The most significant issue is that neither railroad provided the underlying TGMS measurement datasets required by Condition 1(h). Instead, both submissions rely largely on summarized outputs such as movement logs and exception reports. Without the full geometry measurement data, including gauge, alignment, cross-level, warp, and profile readings, it is not possible to verify system performance or determine whether defects were detectable at the time of inspection. In practical terms, the submissions ask FRA to accept the railroads’ conclusions without giving reviewers the data needed to test those conclusions.
Both submissions also raise major coverage and completeness concerns. The available records indicate that TGMS activity was often performed in short or fragmented segments rather than in clear, continuous subdivision-level runs. That makes it difficult to determine whether the full territories proposed for reduced visual inspection were actually inspected during the baseline period required by the waiver. In Union Pacific’s submission, the problem is especially clear because no reporting group achieved full inspection coverage for the relevant period. If baseline inspection coverage cannot be confirmed, then the record does not establish that the waiver conditions supporting reduced visual inspection frequency have been met.
Union Pacific’s submission presents an additional data integrity problem because the different data sources do not consistently reconcile with one another. Movement records, exception reports, and inspector-identified defects do not always align cleanly, including instances where defects appear at locations for which no corresponding inspection activity is clearly documented.
BNSF’s submission raises a related but distinct concern regarding transparency and reviewability. While BNSF did provide supporting materials such as timetables and track charts to the Federal Railroad Administration, those materials were not made available on the public docket prior to implementation due to a technical posting issue. As a result, stakeholders were not able to review or evaluate the full scope of the proposed waiver territory before visual inspection reductions took effect.
Regardless of the reason for the omission, the absence of these materials from the public record at the time of implementation limited the ability of external reviewers to assess the physical and operational characteristics of the territory, including track configuration, traffic conditions, and potential risk factors. This raises broader concerns about whether the waiver is being implemented with a sufficiently transparent and reviewable evidentiary record at the point decisions are made. In addition, portions of BNSF’s submission remain redacted, including information related to ATI platform deployment, further limiting the ability to evaluate how specific inspection systems are being applied across the waiver territory. The data submitted by both railroads also raises questions about whether ATI is producing the claimed safety benefit.
Data Submission Analysis
The disparity between human and automated defect identification is reflected clearly in the railroads’ own data:
BNSF data shows:
Union Pacific data shows:
These figures demonstrate a substantial disparity between defects identified through human inspection and those identified by automated systems. The records also show repeated detections of the same defects over multiple inspection cycles, as well as cases where track inspectors identified defects days or weeks after TGMS had already inspected the same territory. That does not prove ATI has no value, but it does show that the current submissions do not provide a clear, data-supported demonstration that automated inspection is consistently delivering early detection or timely corrective action.
Both railroads also use broad grouping structures that can obscure rather than clarify performance. By combining subdivisions with materially different operating conditions, traffic density, and tonnage levels, the submissions risk averaging performance across dissimilar territories in ways that can dilute defect rates and mask localized problem areas. That makes the data less useful for determining whether automated inspection is performing adequately in the specific places where visual inspections are being reduced.
Taken together, these issues, including the disparity between defects identified by human inspectors and automated systems, the repeated detection of unresolved conditions across inspection cycles, limitations in early detection, and the use of broad grouping structures, highlight a broader concern: automation is being deployed in place of human inspection without sufficient transparency, complete data, or a clearly defined framework for validating safety equivalency. The core issue for Congress and FRA is straightforward: these initial submissions do not provide a transparent, independently verifiable data record sufficient to support confidence in reduced human inspection. The problem is not merely that the railroads have produced unfavorable results. It is that the submissions omit the very data needed to determine whether the results are trustworthy in the first place.
Secure Tracks Act
This legislation would codify the twice weekly visual track inspection requirement that has long served as a core rail safety safeguard, while reinforcing the principle that inspection technology should supplement, not replace trained human inspectors unless and until safety equivalency is clearly demonstrated. (BMWED Applauds Bipartisan Secure Tracks Act Legislation)
Rail safety is inseparable from community safety. Reductions in visual track inspections increase the risk of failures that can impact first responders, surrounding communities, and critical supply chains. As such, it is essential that any transition toward automated inspection systems be grounded in complete data, rigorous validation, and transparent oversight.
Until FRA requires complete TGMS measurement datasets, verifiable proof of full baseline coverage, and reporting that can be reconciled across movement, exception, and defect records, reduced visual inspections are being implemented without an adequate evidentiary basis.
Take action! Use the links below to quickly contact your Members of Congress and urge support for the Secure Tracks Act in the House (H.R. 7784) and the Senate (S. 3987). Additionally, we encourage you to contact your representatives directly by email, phone, or mail. Let’s make sure our voices are heard—pass the Secure Tracks Act. (BILL TEXT)
ACTION NETWORK: SECURE TRACKS ACT - Senate Bill (H.R. 7784)
ACTION NETWORK: SECURE TRACKS ACT - House Bill (H.R. 7784)
Documents
Multimedia
The Brotherhood of Maintenance of Way Employes Division (BMWED) is a national union representing the workers who build and maintain the tracks, bridges, buildings and other structures owned and operated by railroads across the United States. The BMWED is a member of the Rail Conference, International Brotherhood of Teamsters.