MetaRE Inc.-Arup Team Selected as ‘Ideas Challenge’ Semi-Finalist
The Ideas Challenge—short for the Advanced Research Projects Agency–Infrastructure (ARPA-I) Ideas and Innovation Challenge—was announced in August and will award winners a total of $1 million in prizes across two stages.
For Stage 1 of this challenge, ARPA-I received 448 concept paper submissions, according to the USDOT. From those, 15 winners were selected and announced Dec. 2 as semi-finalists (see list below), who will be awarded $20,000 each ($300,000 total) and invited to present and further refine their ideas with business and government stakeholders at the USDOT Innovation Workshop. The Innovation Workshop will be held Dec. 9, 2025, at USDOT Headquarters in Washington, D.C.
In Stage 2, semi-finalists from Stage 1 will be eligible to submit a detailed proposal and up to 10 will be selected as finalists to advance to the ARPA-I Ideas Challenge Finals in 2026, according to the USDOT. The finalists will present their project proposal to a panel of judges and audience members from the public and private sectors to compete for Stage 2 prizes that total $700,000.
The objective of the MetaRE Inc.-Arup project, “Next-gen Smart Sensors Turn Every Train into an Inspector,” is to “demonstrate the viability of novel technologies to detect rails [track] in unacceptable condition at full track speeds,” according to the companies’ one-page summary document. “We estimate that rail operators in the US pay at least $1B per year to inspect rail, logging 10 million inspection-miles per year at a cost of $100 per inspection-mile. Even using today’s fastest platforms for continuous testing, it takes over 19 train-days to inspect a transcontinental corridor, which a revenue service train can travel in less than half that time. We propose to deliver a system to significantly reduce the cost and increase the speed of performing visual inspections of rail.” The system, they said, is “an autonomous, non-contact inspection system based on light and millimeter waves to detect unacceptable wear and external flaws in rail,” and “can be mounted to revenue service equipment and operated at track speeds, allowing inspections twice as fast as currently possible and without any disruption to rail operations.” The team estimates that its system “could reduce the cost per inspection-mile by up to 99%.”
“By providing near-continuous monitoring of any track in revenue service, our system of small, low-cost train-mounted sensors could support remote and AI-enabled rail inspections, the development of digital twins, and robust preventative maintenance programs,” MetaRE Inc. and Arup reported. “The system would also neatly complement traditional ultrasonic inspections for internal defects. It would outperform expensive laser monitoring systems and data-heavy digital imagery systems on cost, capability, ease of implementation, and upside for further innovation. The system uses two subsystems in tandem, one based on millimeter waves (mm-waves) and the other on visible light waves, to detect rail wear and defects in real time at full track speeds. The mm-wave subsystem launches a mm-wave towards the rail to monitor its transverse profile. It captures the reflected waves and converts them into a digital ‘barcode’ that is characteristic of the rail transverse profile. A string of ‘barcodes’ thus generated in real time as the train travels is much less ‘data-heavy’ than other formats, such as high-resolution digital images. This allows for instantaneous diagnosis of the segment of rail in view by a simple comparison between the generated barcode with the one that corresponds to an acceptable rail transverse profile. The mm-wave subsystem is synchronized to GPS to log the exact location of a defect or excessive wear. A fault detection event immediately switches on the optical subsystem further back on the train, which operates as a depth imager to produce an accurate 3D reconstruction of the worn or defective section of the rail. The identified fault can then be verified and classified by a remote observer, either human or AI.”
The MetaRE Inc.-Arup project team includes Nanfang Yu, a Columbia University professor and an inventor of nanophotonic technologies, and Adam Jaffe, a senior engineer at Arup with a background in nanoscience and a focus on commercializing nanotechnologies with applications in the built environment.
