What Bogotá got right.

Bogotá did not invent the bus, but it did insist that a bus could be treated as serious transit. When TransMilenio began service in December 2000, under Mayor Enrique Peñalosa’s administration, it offered a different answer to the city’s daily movement: build dedicated corridors, organize buses into a network, and make the station—not the curb—the place where a rapid service begins.

I am Ines Halvorsen, a former transit planner. The thing I admire about TransMilenio is not a claim that buses are always better than trains. They are not. It is the system’s refusal to let the word bus mean slow, improvised, and second-best. A bus can move quickly and carry many people when the street, stations, payment, service pattern, and operating plan all work together.

The system is often described as Bus Rapid Transit, or BRT. That label can make it sound like a paint job: put a bus in a lane, add a sign, declare victory. Bogotá’s version was more complete. Its trunk corridors use center-running busways; enclosed stations put fare payment before boarding; platforms meet bus floors for level entry; express and local services share corridors; and feeder buses connect neighborhoods to the trunk lines.

What the system got right.

Start with the street. A bus in a dedicated center lane is not waiting behind a line of cars, delivery vans, or a driver who has stopped to ask whether this is the right turn. The lane gives the trunk service a path that general traffic cannot casually take over. Center-running also makes bus priority legible: this space belongs to the service carrying passengers through the corridor.

Then look at the station. Passengers pay before entering the platform area, rather than paying one by one at the bus door. Level boarding removes the step up from street to vehicle and makes it easier for a crowd to get on and off. Doors at the platform and on the buses coordinate the exchange. It is a small piece of choreography repeated across the route, and it saves time without asking riders to hurry.

That station design changes the bus itself. Instead of stopping wherever someone raises a hand, trunk buses serve defined stations. Enclosed platforms can organize queues and protect fare collection; multiple doors can let people leave and board at the same time. The operation begins to resemble rail transit even though the vehicle runs on rubber tires. The distinction matters less to a passenger than whether the trip is clear, accessible, and dependable.

BRT is not a bus with a special name. It is a whole corridor designed to let the bus keep its promise.

The service pattern adds another layer. Local buses stop at each station; express buses skip selected stations. Where the infrastructure includes passing lanes at stations, express services can move around a local bus that is boarding. Riders can transfer between service patterns inside the paid area instead of stepping back to the street and starting over. A faster trip is made by the network design, not by asking one bus to perform every task.

Feeder buses matter just as much as the trunk. A fast corridor that people cannot reach is a fast corridor for someone else. Feeder routes bring passengers from neighborhoods beyond the main busway and connect them at terminals or stations. The trunk carries the long, concentrated part of the trip; the feeder handles the local collection and distribution. Together, they create a larger network than either mode could provide alone.

A rail-like service, built in stages.

The first TransMilenio service opened on December 18, 2000. Its early corridors used dedicated bus infrastructure along major urban roads, with stations, terminals, and connecting routes. The project could be constructed in stages and used the city’s existing street network as its foundation. That does not mean there was no serious construction: busways, station structures, pedestrian access, and operating systems all required design and investment.

The appeal was practical. A bus-based corridor can often be built faster and at a fraction of the cost of a fully grade-separated rail line, while still offering high capacity when the buses, stations, and lanes are designed for the task. The comparison depends on local conditions, land, construction, service quality, and the capacity being sought. The valuable lesson is not a universal price tag; it is that a city has more than one way to build rapid transit.

Speed of delivery matters because a planned system that takes generations to reach riders is not an operating system. Bogotá could open an initial service and extend it over time. The visible infrastructure made the change tangible: a separated lane, a station with clear entrances, a bus that arrived at a platform instead of wrestling for curb space. The network was not complete on opening day, but its central operating idea could be experienced.

The fact that the vehicle is a bus also gives the network some flexibility. Buses can use different service patterns and can connect to streets that do not have rail. That flexibility is most useful when the trunk has strong priority and the stations support rapid boarding. Without those pieces, the bus is flexible mainly in the sense that it can be delayed anywhere.

Where the system struggled.

A successful corridor attracts riders, and success can expose a hard limit. TransMilenio has faced chronic overcrowding: busy stations can fill, platforms can become difficult to move through, and buses can arrive already packed. A dedicated lane helps buses move through traffic, but it does not automatically expand the number of people a station can safely process or create room inside a vehicle at the moment people need to board.

That is why capacity has to be thought of as a chain. The busway, the station platforms, the entrance and fare gates, the number of doors, the service pattern, and the surrounding walking routes all shape how many people can use the corridor comfortably. If one link is undersized, adding buses may simply bring more vehicles to a crowded station. Express and local operations can improve the use of a corridor, but they also require clear wayfinding and careful management.

Maintenance is another unglamorous test. Stations, pavement, vehicles, fare equipment, elevators, lighting, and doors all wear out. When maintenance is deferred, the problem does not stay on a work order: it reaches the platform as a broken escalator, a rough surface, a closed entrance, or a less reliable trip. A system that moves a great many people needs a steady plan to renew the things it depends on, not only a ribbon-cutting plan for the next extension.

Bogotá’s long-running debate over buses and a metro is not a simple contest with a final winner. The city has argued for decades about the role, cost, route, and timing of rail transit alongside a bus-based network. The first line of Bogotá’s metro is now under construction. That decision does not erase TransMilenio’s work, nor does the existence of a BRT network make the metro unnecessary. A city can need more than one kind of high-capacity service.

The difficult question is how the systems will meet. The metro will need useful connections to buses, safe and legible transfers, fares that work across services, and streets that let people reach stations. TransMilenio also needs continued attention as a major part of daily mobility, not a temporary bridge that can be neglected now that rail construction is visible. A network is only as connected as the trip a passenger can actually make between its parts.

What the BRT label can hide.

In North America, a project can be called BRT when it has a branded bus, a few upgraded stops, or a limited stretch of bus lane. Those can be useful improvements. They are not automatically the same as a rapid-transit corridor. The word on the sign cannot make a bus reliable if it spends most of its journey in mixed traffic or loses its right of way at the intersection where delay begins.

Dedicated lanes are the first promise to keep. They need to be continuous enough to matter, enforced enough to remain clear, and designed through the places where congestion is worst. A short lane that disappears before a bottleneck may help, but it cannot deliver the same running-time consistency as a corridor where the bus retains priority through the pinch point.

Off-board payment and level boarding are not decorative upgrades. They move fare decisions away from the doorway and reduce the pause each time a bus stops. Level platforms and multiple doors make the movement of riders easier and more accessible. Together, they turn each station stop into a quick exchange rather than a queue at the front step.

Frequent service is part of the product too. A well-designed station with long waits is still a long wait. Trunk and feeder schedules need to connect, and the buses need to show up at useful intervals throughout the day. Frequency without priority can bunch in traffic; priority without enough service can leave the corridor underused. The pieces make sense together.

• Protect a continuous right of way with dedicated bus lanes, especially through recurring bottlenecks.
• Collect fares before boarding and provide level, multi-door access at stations.
• Run frequent trunk services with express and local patterns only where the corridor and stations can support them.
• Connect feeder routes to the trunk and make transfers simple to understand.
• Fund maintenance and operations alongside construction, so the service keeps working after opening day.

The lesson for cities that want rapid transit.

I would not tell every American city to copy Bogotá line for line. Street widths differ. Development patterns differ. A corridor that works in a dense Latin American capital may need a different station or vehicle arrangement in a smaller North American city. The useful question is not, “Can we import the same map?” It is, “Which features create the reliable, high-capacity service we are promising, and are we willing to build them?”

Start with the corridor’s job. Is it moving people through the center, connecting neighborhoods to a rail station, or linking suburban centers? Then choose the service pattern, lane design, station form, and frequency to match. If the bus is meant to carry the main load, protect its path. If passengers need to transfer, design the transfer area. If fares slow boarding, change where and how payment happens.

The budget conversation should include the full life of the system. Construction attracts attention, but operators, vehicles, station cleaning, security, pavement repair, fare equipment, and service frequency determine the everyday experience. A cheaper project can become expensive if the city builds infrastructure it cannot maintain. A more modest project can be transformative if it reliably connects useful places and has the support to keep running.

For planners, one of Bogotá’s most important lessons is institutional. A trunk-and-feeder network needs someone responsible for the whole service, not just a lane contractor and a set of unrelated route operators. The public agency must plan, coordinate, monitor, and adjust. When a bus misses its connection or a station cannot handle the crowd, there has to be a clear owner of the problem and a way to correct it.

For riders, the test is wonderfully ordinary: Can I get to the station, understand where to wait, board without a long bottleneck, and trust the next part of my trip? Bogotá showed that buses can form the spine of a rapid-transit network when the city gives them infrastructure and operations equal to the ambition. It also showed that opening a system is the beginning of stewardship, not the end of planning.

When I see a “BRT” proposal now, I look past the acronym. I look for the bus lane that will still exist next year, the fare reader that is not at the front door, the platform that meets the vehicle, the service that comes often, and the feeder connection that gets someone to the trunk. If those details are there, the label may mean something. If not, it is just lettering on a bus.