Building a safer future with trustworthy AI

As Nenad Mijatovic, our Chief AI & Data Scientist, explains, "I remember the first time I presented on Trustworthy AI to the Communications team some years back. Amidst all the excitement around data science and AI's transformative power, they felt my topic was a bit…stiff. But we landed on it being essential. If you want to lead change in rail, where trust is paramount for daily commutes and efficient, low-carbon logistics, you must develop AI-powered solutions that are not just innovative, but also trustworthy." 

This commitment has been our guiding light at Alstom. Nenad adds, "I believe the biggest opportunity lies in AI's ability to understand specific environments, like complex rail networks, and make explainable and safe decisions – grounded in expert-driven learning and quality data, with strong guardrails in place."

Pierre Eisenmann, our Data and AI Chief Architect, highlights the transformative power of Data and AI:  "We empower employees and clients, from engineers to rail operators, by giving them access to actionable insight on how our systems work, through rich data from operations and AI analysis. AI tools can be used for predictive or corrective maintenance, like our radio failure prediction system, which avoids train communication interruptions by resolving issues before they affect rail operations. This improves transportation reliability and builds trust in AI when people understand its processes and observe reliable decisions." 

Robust radio communication is vital for millions of commuters and freight operators, a significant interruption can cause delays and unwanted cancellations. Alstom’s innovative AI tool for predicting radio failures in rail signalling was presented at the Welcoming event by make.org and chosen from over 550 submissions as one of the 111 projects for the "AI for Efficiency Call for Projects" by the French Ministry of Economy and will be featured on the AI Summit Observatory platform.

Pablo Celada-Alonso, AI Solutions Deployment Director, highlights the practical application of our commitment. "This pledge, while significant, boils down to the details. For example, our 'AI Lab for All' democratises AI access for Alstom employees. The objective is to make AI tools, resources, and knowledge available to a wider audience, now our employees interested in exploring AI capabilities in a secure, trusted, and agile environment will now have the necessary support and access."

"We share a common ground here," says Kenza Saiah, our Data Science and AI Manager. "With AI grounded in expert-driven learning and quality data, and the right guardrails in place, the possibilities in the rail industry are endless. From conserving energy to optimizing schedules and predicting maintenance, we're enhancing the passenger experience like never before."

At Alstom, we are driven by a vision to create mobility solutions that offer enjoyable travel experiences and drive societies towards a low-carbon future. This vision requires a delicate balance of innovation, robust processes, and thoughtful regulation. For us, data science and AI are actively shaping our industry, from enhancing passenger safety to optimising train operations across complex rail networks. Our journey towards Trustworthy AI is a continuous one, and signing this pledge reinforces our commitment to leading the way in developing AI-powered mobility solutions that are not only smart but also safe, dependable, and ethical.