The autonomous enterprise: Transforming how public transport operates

Discover how the Autonomous Enterprise helps public transport operators reduce costs, improve reliability, and deliver better passenger experiences with AI.

If you lead operations, assets, or planning at a public transport organization, you already know the pressure. Ageing fleets demand more maintenance than budgets allow. Ridership patterns shift faster than schedules can adapt. Many existing systems already use historical and real-time data to monitor operations, optimize activities, and support predictive planning. However, these capabilities often remain fragmented across functions, and translating insights into coordinated action can still require significant manual effort.

That gap creates an opportunity to connect data, decisions, and execution more effectively across the operation.

An Autonomous Enterprise for public transport connects business data, applications, AI assistants, and agents to support decisions and automate tasks across operations and services. AI cannot only surface insights and recommendations but also execute routine actions and workflows within defined controls. This helps employees act earlier and focus on situations that require experience, accountability, and human judgement.

Why does public transport need a more connected operating model now?

Public transport has always operated at the intersection of public accountability, financial constraints, and operational pressures. Today, that tension is sharper than ever. Ridership recovery is uneven across networks. Workforce challenges are intensifying. In the United States, public transit agencies delivered 7.7 billion passenger trips in 2024, yet ridership was still only 85% of pre-pandemic levels, according to the American Public Transportation Association. Workforce capacity remains another pressure point. In a 2024 survey of 117 U.S. public transit agencies, 85% reported continuing worker shortages, with bus operator roles particularly difficult to fill.

At the same time, asset pressure is mounting. Buses, metros, trams, trains, and supporting infrastructure must perform reliably across every service, yet maintenance decisions are often made from incomplete data spread across separate systems. The scale of the asset challenge is significant. The U.S. Federal Transit Administration estimated that $140.2 billion, or 10% of U.S. transit assets by replacement value, were not in a state of good repair in 2022.

Sustainability obligations add another layer: operators are expected to reduce carbon dioxide (CO₂) emissions per passenger kilometer while managing significant cost pressures.  In the European Union, road transport accounted for 73% of transport greenhouse gas emissions in 2023, reinforcing the importance of more efficient passenger mobility and public transport.

Compounding these pressures is a patchwork of planning tools, finance systems, and operational applications that don't connect, forcing teams to assemble data by hand and respond to events after they've already disrupted service.

Public transport will not become more resilient by automating isolated activities. The greater opportunity is to connect decisions across assets, workforce, network operations, finance, and passenger services.

What does 'autonomous enterprise' actually mean for public transport?

The term 'autonomous' is used broadly in transport, and that creates confusion. An Autonomous Enterprise isn't about removing people from business decisions. It's about changing what people spend their time on and improving overall effectiveness.

Consider what can happen when a vehicle's fault is detected overnight. Many public transport organizations already use automation to identify issues and coordinate responses quickly. But where information or workflows remain fragmented across maintenance, parts, workforce, and service-planning systems, resolving the issue may still require additional manual coordination. An Autonomous Enterprise can take existing automation further by enabling AI Assistants and Agents to connect these processes, cross-reference maintenance history and parts availability, identify qualified technicians, and surface or execute appropriate next steps within defined controls.

The Autonomous Enterprise is an operating model that enables AI to transform how people work and how processes are run by grounding decisions in real-time intelligence, automating end-to-end workflows within governed boundaries, and proactively improving operations to deliver better outcomes for their passengers and communities.

AI can detect problems and propose potential solutions. People use those insights to make informed decisions, and AI can then execute approved actions within defined controls. This represents a structural shift in how the enterprise operates, enabled by AI, connected data, and coordinated workflows.

For public transport organizations, this shift can extend across strategic planning, passenger services, workforce management, network operations, asset management, and finance.

Why public transport is well-suited to AI

Public transport operations are, at their core, high-volume, asset-intensive, and tightly regulated. That combination makes them particularly well-suited to the kind of continuous, AI-enabled optimization that the Autonomous Enterprise enables. Consider where autonomous agents can directly impact outcomes:

• Predictive asset maintenance: Agents monitor real-time asset health data, cross-reference maintenance history and parts availability, and proactively schedule interventions, reducing unplanned downtime and shifting maintenance from reactive to condition based.
• Network operations and scheduling: When passenger demand shifts or a service disruption occurs, agents can model alternative scenarios across the network, recommend adjustments, and trigger the necessary operational responses in real time, without waiting for a planner to convene.
• Workforce planning and scheduling: Agents match crew availability, qualifications, and regulatory rest requirements against dynamic service needs, helping operations teams fill shifts efficiently and reduce the administrative burden of manual rostering.
• Passenger experience and demand management: AI-driven analytics can surface demand patterns, support dynamic pricing strategies, and enable personalized communications, helping operators improve farebox revenue and passenger satisfaction simultaneously.
• Sustainability reporting and compliance: Agents can continuously calculate CO₂ per passenger kilometer, align grant and fund reporting with sustainability priorities, and surface insights that help operators meet regulatory obligations without manual data assembly.
• Finance and strategic planning: Integrated financial intelligence enables scenario-based profitability analysis, so leaders can evaluate route investments, capital programs, and fleet decisions with a clear view of financial and operational consequences.

Each of these represents a real operational pain point for operations managers, asset teams, and workforce planners across bus, metro, light rail, and regional train networks. The opportunity isn't to automate for automation's sake; it's to free experienced people from high-volume routine work so they can focus on decisions that genuinely require human judgement.

What makes this different from earlier AI approaches?

Many public transport organizations have already experimented with AI, a predictive maintenance tool here, and a demand forecasting model there. The consistent frustration is that these tools improve one corner of the operation while leaving the bigger picture unchanged. They optimize silos.

The Autonomous Enterprise is built on three foundations that matter for enterprise-scale deployment:

First: Agents don't just read data; they reason within the context of how the business runs, including the workflows that matter most in public transport: maintenance planning, workforce scheduling and finance.

Second: AI assistants and agents can see the relationship between a vehicle's fault, an open maintenance work order, crew availability, a service schedule, and a finance position, simultaneously. That connected view is what enables genuine end-to-end action, not just isolated recommendations.

Third: Enterprise-grade governance. Every AI action needs to be auditable and traceable. For public transport operators working under public accountability requirements, service performance obligations, and sustainability regulations, this isn't optional.

Where should public transport leaders start?

Public transport leaders can start by identifying where the gap between a signal and a coordinated action creates the greatest operational impact. The most useful question isn’t “How much AI should we use?” but “Where can greater autonomy help us sense, decide, and act more effectively?”

That might be the time between detecting a vehicle's fault and coordinating a maintenance response. It might be the lag between a shift in passenger demand and a service adjustment that reaches passengers. Or it might be the hours spent assembling sustainability, workforce, or financial data from separate systems before a decision can be made.

For public transport leaders, the opportunity is to progress toward a more AI-assisted operating model by connecting data and workflows across the areas where delayed information or manual coordination has the greatest operational impact. The goal is not to automate everything at once, but to expand autonomy deliberately, with people retaining oversight while AI supports decisions and executes appropriate work within defined controls.

Public transport organizations have a fundamental responsibility to move people safely and reliably while maintaining service levels and making effective use of limited workforce, assets, and infrastructure. Progressing toward a more AI-assisted operating model can help leaders coordinate these resources more effectively, respond to disruptions earlier, improve operational resilience, and support sustainability goals while delivering better outcomes for passengers and communities.