Every public transit agency has a strategic plan, and it's genuinely ambitious: grow ridership, close the workforce gap with a stronger team, win the next round of public funding, hit an electrification target, build a service people trust enough to plan their lives around. Those are big, exciting goals, worth getting right. The best agencies are already asking a sharper question about how to get there faster: what would it take for the team we have to cover more strategic ground than anyone thought a team our size could?
That's exactly where agentic AI earns a seat at the strategy table. Our ebook, Agentic AI for Public Transportation, makes the operational case: what an agent actually does across planning, scheduling, operations, and control, and why. This is the case one level up: what it makes possible for an agency that's serious about where it wants to be in five years.
Nearly every agency strategic plan right now has a line about the workforce shortage, because it's real, global, and not closing on its own. A 2025 joint report from the International Transport Workers' Federation and UITP found a 2.4 million-person shortfall in urban public transport jobs worldwide, even after a decade of 20% workforce growth, across markets from the US and Canada to Germany, France, the UK, Japan, and Australia. The instinct is to treat that as a hiring and retention problem. It's also a capacity problem, and capacity is something an agency has more control over than the labor market.
When a junior scheduler can operate with the reasoning of a senior one behind them, and a senior planner spends their week on network design with the shift corrections handled for them, an agency effectively gets more capacity out of the team it already has. That gives a hiring strategy real room to work on its own timeline, backed by a team that's already covering more ground than its headcount would suggest.
Grant and subsidy cycles increasingly reward agencies that can demonstrate rigor: clean data, fast turnaround on compliance reporting, a credible story about how service decisions get made. In the US, that shift is visible in the Department of Transportation's own move to agentic AI for grant-compliance review, used to verify applications and flag missing information so federal dollars move faster.
The European Union's transport funding programmes are moving in the same direction, tying billions in Connecting Europe Facility funding to the kind of digital reporting and data-sharing standards agencies now have to meet just to qualify. Wherever the funding sits, the agencies applying are being evaluated by a process that increasingly expects the same clarity and speed they're now capable of producing internally.
An agency that can turn "show us how this schedule change affects coverage and cost" into a same-day, presentation-ready answer walks into that conversation with a stronger hand, whichever government is writing the check.
Riders come back because the bus was there, and it was there again the next day, and the day after a storm rerouted three lines it still mostly worked. Reliability is the actual lever behind ridership growth, and the pattern holds wherever you look. Research on Miami-Dade's bus network found that as ridership recovered after the pandemic, on-time performance became an increasingly strong driver of that recovery, paying off most on the routes people already rode the most. The UK's official passenger watchdog found that nearly half of bus passengers say the bus is their only real way to make a given trip.
Agentic AI's clearest contribution here may seem small and unglamorous: it shrinks the time between a disruption happening and a rider actually being told about it, correctly, everywhere they'd look. That's a small technical detail with an outsized effect on whether people trust the service enough to plan their day around it.
The agencies getting real, compounding value out of agentic AI are the ones where a director or a chief planning officer decided this was worth owning at the strategic level, with the reasoning and standards set from the top. That's the difference between a pilot that stays a pilot and a capability that becomes part of how the agency runs.
APTA's 2026 survey of North American member agencies found that AI adoption today is concentrated in customer support and analytics, with back office and operations, the areas closest to planning and scheduling, still mostly on the roadmap rather than in production. UITP is running the same conversation on the other side of the world: its members are convening in Delhi in October 2026 specifically to work through where AI belongs in daily transit operations. Wherever an agency sits, the ground closest to planning and scheduling is still mostly open, for whoever moves there first.
That's also the bet we made at Optibus. Optibus Agent was built because we believe the agencies that treat agentic AI as core to how they plan and operate are the ones who'll be measurably ahead of the ones who waited to see how it played out elsewhere first.
Good transit still runs on people who understand their riders, their drivers, and their city. What agentic AI changes is how much strategic ground an agency can actually cover with the people it already has, which is exactly the constraint most strategic plans are quietly built around.
Further Reading