Where Transit Technology Is Heading: From Devices to Decisions

MAGNUS FRIBERG
Chief Executive Officer
Luminator Technology Group

In under a year, AI has moved from a productivity tool that helps employees work faster to one that is changing how entire industries operate. That shift is already happening in public transit, and this fall at APTA’s EXPO in Chicago, the conversation will take a major step forward.

Consider the pace elsewhere: 90 percent of technology professionals now report using AI at work. In emergency medicine, ambient AI scribes have cut physician documentation time by 28 percent per encounter, time returned to patients rather than paperwork. In professional services, AI is pulling firms away from headcount and billable-hour economics toward recurring, platform-based revenue. The common thread is that AI is not just improving products at the margins. It is changing how value is created, and transit is not exempt.

The question for our industry is whether we lead this change or follow it. Today, we sit closer to the starting line than many realize. In a May 2026 APTA survey of 32 member agencies, at most, half were even considering AI in any single operational domain, and only 16 percent for fares and ticketing. That is not a criticism, it is an opening; and given how quickly AI adoption is accelerating, those figures have likely grown since.

The early proof is already global, and it is operational rather than theoretical. In New York, MTA’s AI-assisted bus-lane enforcement has helped lift bus speeds by five percent and reduce collisions by 20 percent on the routes where it runs. In Barcelona, an AI ventilation system in the metro cut energy use by roughly a quarter while improving passenger comfort. Across Europe, more than 80 percent of major cities have adopted AI-driven ticketing, and London’s biometric gates have shortened boarding times by 40 percent. Add driver-efficiency fuel savings at Alsa in Morocco, smart charging at Arriva in Spain, and the field research gathered by the Urban Institute, the ITF at the OECD, and Arup, and a pattern emerges. The agencies moving first are not buying novelty. They are buying outcomes.

That distinction matters, because the real change is not any single tool. Modern transit has quietly become a digital ecosystem: destination signs, onboard cameras, passenger counters, the operations software that runs the fleet, the data all of it generates, and the intelligence that interprets it, with skilled people setting the priorities and making the judgment calls. The value, and the disruption, live in the connections between those parts, not in the boxes themselves. An arrival prediction is only as good as the vehicle-location data behind it. A safety alert is only as useful as the workflow that acts on it.

Seen this way, what AI means for agencies is concrete. It means anticipating a component failure before it becomes downtime, so vehicles stay in service and assets last longer. It means turning streams of data into decisions an operator can act on in the moment, rather than dashboards no one has time to read. And it means earning rider trust the only way it is ever earned, through service that holds up day after day. The result is lower operating cost, longer asset life, more efficient labor, and riders who choose to come back. Agencies that treat AI as a capability woven through the whole ecosystem, rather than a feature bolted on to trim a line item, will be the ones that lead.

I write this from a particular vantage point: Luminator has spent nearly a century in this industry and today serves operators in more than 85 countries, working across software, hardware, data, and AI-driven insight rather than any one layer. That breadth is why we established a dedicated Architecture and Innovation function to define how our systems fit together from end to end, to build the data architecture that unlocks system-wide insight, and to guide responsible AI adoption.

Intelligence can live at every step of that chain: video that understands what a camera sees, software that acts on it, and analytics that turn it into foresight. My conviction is straightforward. The most credible voice on what AI means for transit is not the one with the newest single product, but the one that can see the whole picture.

So, what does this mean for the agency leader weighing where to invest next? Think in terms of your technology stack: the layers of devices, software, data, and analytics that will increasingly run AI at every step. The advantage does not come from adding intelligence in one place; it compounds when the layers are planned to work together, so a smarter camera feeds smarter software that sharpens the analytics guiding the next decision. The practical takeaway is to treat your next procurement less as a shopping list of features and more as an architecture, and to arrive at EXPO with that question in hand: not which product is newest, but how do the pieces connect to move your numbers on uptime, cost, and rider satisfaction.

The choice ahead is not whether AI arrives in transit. It is whether we meet it as a scattered collection of tools or as one connected ecosystem built around the rider. When our industry convenes in Chicago this fall for APTA’s EXPO, that is the conversation worth having, and the moment to begin having it together. The future of mobility will be shaped by those who arrive ready to lead it.