Ferrovial puts AI at the heart of its transformation

With the aim to accelerate innovation and transform its digital infrastructures and services, Ferrovial created its Digital Hub to serve as a meeting point where research and experimentation with digital strategies could, for example, provide new sources of income and improve company operations.

Since its creation over five years ago, the Digital Hub has included a team of experts in innovation, technologies, and trends — such as IoT, big data, AI, drones, 3D printing, or advances in customer experience — who work in concert with other business units to identify and execute new opportunities. There’s no defined hierarchical structure, but the Hub is based on a transversal and collaborative model where groups are created, and employees from all areas work on specific projects along with external experts.

One of the main strategic transversal areas is AI. According to Luis Carlos Prieto Fernández, head of Ferrovial’s AI center of excellence (AI CoE), its value is demonstrated by the work the company has been doing with all business units. “They have AI-based services that either make processes automatic, more efficient, or provide the business with services that give competitive advantages,” he says.

With this in mind, the AI ​​CoE acts along two clear lines at a global level. The first is the development of AI models and projects for business units that create digital solutions together with the rest of the Digital Hub team. The second, given Ferrovial’s global nature, is its continuous contact with all units, acting as an AI knowledge hub. “This allows us to contribute ideas and solutions we’ve already been working on, and avoid duplicating proofs of concept and projects,” he adds.

Getting AI off the ground

After several years of innovating in AI, Ferrovial can cite many examples of projects and services benefitting from it. “If we look at what we could call classic AI, an example would be our Herbicide Train project, where front cameras have been installed in several trains throughout Spain that don’t go below 50 km/h,” Prieto says. “And combined with the help of our computer vision models, we’re able to detect when there’s vegetation between the rails and shoot herbicide up to seven cars behind.” Being able to do this has economic as well as ecological impacts — reduction in the cost of glyphosate by more than 80%, regarding the former, and only irrigating where necessary, for the latter.

Plus, to provide all its businesses with the ability to obtain information in record time, the company has been working with gen AI and LLMs, specifically Azure OpenAI, since December 2022. So projects to analyze huge amounts of data in multiple formats already in production are now considerably more efficient and cost effective.