Measuring What Matters
October 1, 2026
In January, I wrote that Aii would return to first principles in 2026. That meant leading with innovation, staying nonpartisan, and judging ideas by their results. Nine months later, it is time to report on that work.
Rail safety: counting versus consequence
In June 2025, I testified before the House Subcommittee on Railroads, Pipelines, and Hazardous Materials that rail safety rules should rest on the data and science. Since then, the Federal Railroad Administration has approved a national waiver for automated track inspection. This week Aii followed up with data. Our new report, Man/Power, reviews 25 years of FRA accident records.
Critics of automated inspection point out that it covers only 6 of 23 inspection categories, about 26 percent. That is accurate; but accident history tells a different story. Track geometry caused 38.8 percent of derailments attributed to track, roadbed, and structure, that’s nearly forty percent. Of those, 96.2 percent involved conditions measurable by track geometry systems. As Aii Executive Director Ben Dierker put it, “Six of 23 tells policymakers how the requirements are counted, not how significant they are in actual derailments.”
With advances in sensors, machine vision, and artificial intelligence, I would not be surprised to see the number of conditions automated systems can detect double within the next five years.
None of this is about losing jobs by the way. The men and women people who work the track to locate track defects are the same people who repair them. When machines find more defects, sooner, those skilled workers spend more of their time fixing track. That’s a win for workers and safety. Visual inspections also remain essential for the conditions technology cannot yet see. The goal is to put people to work where they add the most safety. Read the full report (PDF).
Surface transportation: getting the long-term bill right Congress has extended surface transportation programs through December 11. That buys time, but it is not a plan. As I write the Senate is closing in on has remained a necessary but elusive deal, fixing the timeline for building infrastructure projects. Fingers crossed. In May, Aii wrote to the House Transportation and Infrastructure Committee on the BUILD America 250 Act. In July, we hosted a panel of national experts on the path ahead. Four principles guide our view of how we should move forward:
Fund it honestly. The Highway Trust Fund is structurally insolvent. It depends on repeated transfers from the general fund. Short-term extensions delay the problem without solving it.
Build smarter. Performance-based standards, digital project delivery, and modern materials deliver more value from every dollar. Federal policy should not welcome automation in trucking while resisting it in rail.
Fix permitting. As mentioned above, time itself is a cost. Clearer review procedures and wider use of categorical exclusions protect the public interest without years of delay. Bring in private capital. Gross federal debt passed $40 trillion this September, and interest takes a growing share of every budget. Federal dollars alone will not close the infrastructure gap. Well-structured public-private partnerships, a higher cap on private activity bonds, and flexibility on tolling can bring private investment into public projects and give them accountability over the full life of the asset.
Other recent work
Here’s a list of some of our other recent work that our team has delivered. Each one deserves a close look.
- A New Transcontinental Railroad and What a Coast to Coast Railroad Company Could Mean look at the proposed rail consolidation.
- The Politics of Data Center Uncertainty examines the power and siting questions behind the AI buildout.
- Telosa and the Challenge of Building a City from Scratch asks what new cities can teach existing ones.
- Holding Back Disaster, by Scott Landes and Benjamin Dierker, continues Aii’s long work on excavation safety and damage prevention.
Innovation can me technology but often it also means looking again to see if we’ve missed something. This quarter, that meant looking past how rules are counted to how accidents actually happen. We will keep measuring what matters.
Brigham A. McCown, Chairman