Republic Airways Taps New Collins App For Maintenance Log Insights

ATLANTA—Republic Airways has become the first airline to implement a new application developed by Collins Aerospace to help automatically resolve coding and freehand text errors in aircraft maintenance logs. It will use the technology to generate insights for its fleet of more than 200 Embraer 170 and 175 aircraft.

First released in late 2024, Ascentia Repeaters is an application on Collins’ Ascentia predictive health platform that uses natural language processing (NLP), a type of artificial intelligence that enables computers to understand and communicate with human language. The application digests unstructured data from different sources—such as from technicians, pilots and the aircraft itself—and identifies, classifies and clusters common maintenance issues together to show what Collins says are clear areas of correlation and recommended fixes.

According to Collins, the application helps airlines quickly pinpoint maintenance priorities based on information specific to their fleets, operations and historical data, which should improve final troubleshooting effectiveness, proactively prevent recurring issues and reduce aircraft downtime.

“With nearly 30% of all unscheduled airline events preceded by multiple maintenance reports, Ascentia Repeaters identifies the often time hidden patterns from these write-ups, highlighting common events and allowing airlines to take proactive maintenance actions to prevent future disruptions, improve repair effectiveness and enhance operational reliability,” says Nicole White, vice president and general manager of connected aviation at Collins Aerospace.

According to Matt Suckow, director of enterprise resource planning and maintenance technology at Republic Airways, the airline chose Ascentia Repeaters because it simplifies maintenance by “turning scattered maintenance data into clear, useful insights.” He expects it to help the airline "quickly identify recurring issues, reduce downtime and keep our Embraer E-Jet fleet running reliability.”

Seth Babcock, Collins’ head of connected aviation tech ops solutions and analytics, tells Aviation Week Network that the company built Ascentia Repeaters using log page data sets from about 15 different airlines. “A lot of work was tied into developing that [NLP machine learning] model, developing entity recognition type things around that and then looking into a very specific keyword dictionary that can assist in a lot of NLP things,” he says. “So, for instance, if a log page free text has ‘A/C’ in it, [determining if] the contextual of that is ‘air conditioning’ or ‘aircraft.’”

Babcock adds that Ascentia Repeaters regularly ingests airline data, runs analytics and provides 24/7 alerting that recognizes systemic issues from historical data. He says Collins typically runs analytics on an approximately 60-day data set and the application averages around 8-10 alerts per day.

Beyond leveraging historical data for better maintenance insights, Collins also is developing a new AOG application for Ascentia to track aircraft out of service. It plans to release it in the third quarter.

At MRO Americas here April 8, Collins Aerospace became the fifth member of the Digital Alliance for Aviation, a group jointly developing digital solutions for airline operations powered by the Airbus Skywise platform.

While Ian Galloway, Collins’ senior director of applications and connectivity solutions, says the partnership will not focus on Ascentia data, the company will be providing its predictive maintenance and analytics expertise for components such as air conditioning supplemental cooling, electrical power distribution center, hydraulic power and engine bleed air supply parts.

“There are a lot of problems that we’re not going to be able to solve on our own,” says Galloway, noting that the aviation industry is generating massive amounts of data from around 30,000 commercial aircraft, “so we’re excited to work on some very specific areas with Airbus and the rest of the [Digital] Alliance team.”