African railways developing new networks could avoid some costly conventional inspection systems by integrating drones and artificial intelligence (AI) from the outset, speakers said during a Chartered Institute of Logistics and Transport (CILT) Global Rail Group webinar.
The webinar, the first in the group’s Africa Regional Series, examined the potential of drones and AI to improve railway inspections and the barriers preventing operators from moving beyond proofs of concept and pilot programmes.
Developing railways could limit investment in complex sensor-based systems by incorporating drones and AI into inspection processes while building new networks, said TrackSense by DecisionWorks President and Director of Flight Operations Grant Barkman.
New railways could first determine what drones and AI are able to provide before investing in additional inspection infrastructure, he said in response to a question about the potential benefits for countries such as Ethiopia.
Saudi Arabia Railways has discussed using drones and AI as a starting point for inspecting newly built track before adding other systems, Barkman said.
The technology could be particularly useful in Africa where some countries are relatively new to railway operations or have lost rail knowledge and experience, said CILT Global Rail Group Chair Andrew Young.
AI could help operators interpret inspection data, analyse trends and improve the reliability and safety of railways while reducing the resources traditionally required, he added.
Drones can collect several types of inspection data during a single flight, including track geometry, ballast volumes, vegetation encroachment, sleeper condition, rail gaps and infrastructure defects, according to Barkman.
Comparatively low drone operating costs could enable more frequent inspections and allow maintenance teams to concentrate on high-risk areas, he said.
Drone-based bridge inspections could also be completed without track occupation permits, which could avoid interruptions to revenue-generating rail operations, Barkman added.
AI can automate the analysis of inspection data, identify infrastructure defects and monitor changes in asset condition over time. Its value also lies in integrating information from drones, manual inspections and other sensors into a single defect management system, Barkman said.
However, organisational and workflow barriers impede adoption.
TrackSense has conducted dozens of discovery sessions with railways that recognise the technology’s potential and business case, according to Barkman. He estimated that only about one in 10 engagements have progressed beyond initial interest, proof-of-concept or pilot stages.
“The challenge is less about whether the technology works and more about how it can be integrated into existing railway workflows, organisational structures and personnel responsibilities,” Barkman said.
Successful implementation requires operators to assess their organisational readiness, develop a business case, train personnel and introduce the technology through staged pilot programmes, webinar attendees heard.
Drones and AI could also improve safety for railway employees involved in maintenance and inspections, particularly on networks operating in challenging environments, Young said.
