Planning to Deploy A CBTC Solution? Avoid this Mistake

Transit authorities planning to transition from conventional to CBTC signaling must treat the depot and mainline as a single entity; otherwise the boundary becomes a barrier for launching trains into service. The barrier results from CBTC and conventional signalling speaking different languages; a simplified interface will lose something in translation, preventing a seamless handover of a train from depot to mainline.

Transit agencies planning to deploy a CBTC solution must be mindful that a CBTC solution is effective only when it has control over all aspects that affect mainline operations. The time it takes to launch trains from the depot is also a factor because it compromises the throughput on the mainline. Non-CBTC actors, such as a conventionally signalled depot, hinder a CBTC solution’s ability to control the flow of trains on the mainline, reducing the advantages CBTC was meant to introduce.

Implementing a CBTC solution on the mainline and leaving the depot conventionally signalled is a mistake. Below, I describe three design options: Option 1 is flawed; Option 2 is a compromise; and Option 3 is the gold standard.

Option 1 – Mainline CBTC, Conventional Depot

Figure 1 shows a typical mainline/depot alignment with the hashed lined depicting the boundary that transit agencies define to separate their operations; the depot operator has authority over the depot control area and the mainline operator has authority over the mainline control area.

Transit planners typically use the mainline/depot boundary to separate the CBTC control area from the conventional control area (Figure 2), if the decision is to implement CBTC on the mainline only (depot remains conventional) due to cost or other reasons.

At first glance, this is a simple and elegant arrangement; straight line down the mainline/depot boundary separating CBTC from the conventional signalled territory. However, this is based on a lack of understanding of the characteristics of a CBTC solution and problems emerge when launching trains into service.

Initializing the vehicle controller (VC) is a prerequisite for launching a train. Several initialization steps are performed; however, the critical step is localizing the train by crossing two beacons to gain position (Figure 3a). This enables the CBTC system to “see” the train. These beacons must be placed within the CBTC territory and since the territory is confined to the mainline, it creates an operational nightmare.

Launching trains requires the train to move from the storage lane to the hostler in manual mode following conventional signalling rules (Figure 3b). Since the VC’s are not localized (beacons are located on the mainline), the CBTC system cannot see the train. Before the train can enter the mainline, the CBTC system imposes draconian measures to protect the train by creating a protection zone from the hostler to Station 1. Other trains operating on the mainline are denied entry into this zone.

The train is permitted to move toward Station 1 once the protection zone is created (Figure 3c). Train position is established as it crosses the beacons and localizes (Figure 3d). Only at Station 1 can the train enter into ATC (Automatic Train Control) mode and depart. By this time, trains are bunching behind Station 1, slowing service.

If the train does not localize (beacon antenna has failed) or another type of self-test failed (train door test for example), the train cannot reverse direction back to the yard; the hostler is occupied by the next train waiting to depart for Station 1. The protection zone must be extended past Station 1 to the nearest crossover and circled back to the yard on the bottom track. This is a large swath of track to take out of service during passenger carrying hours.

Under this arrangement the operator is not aware the train is unfit for service until it reaches Station 1, which is too late. The operator is left with no options to remove the train from the mainline other than to continue on to the nearest crossover, causing further disruption to service.

One alternative is to implement fallback mode of operation (See my previous post about Fallback Mode of Operation) but this is a costly capital expense and an over-engineered solution to a simple problem.

Option 2 – Hybrid Configuration

Under the hybrid option, the CBTC boundary is extended to include the hostler, permitting the localization beacons to be placed inside the depot (Figure 4a). The train follows conventional signalling rules as it leaves the storage lane (Figure 4b) and establishes position when crossing the beacons (Figure 4c). The advantage is that by the time an ATC ready train arrives at the hostler it can depart with no disruption to mainline operations (Figure 4d). Further, if the train is not able to localize or another self-test failed, the operator can route the train from the hostler back to the storage lane without ever entering the mainline.

The hybrid option enables the operator to: launch trains in ATC mode from the hostler; identify failed trains before they enter the mainline; and route failed trains from the hostler to the storage lane. These are critical features missing in Option 1.

Option 3 - Gold Standard

The ideal configuration is full CBTC control of the depot (Figure 5). The trains are ATC ready in the storage lanes and localization is not required. At the appointed time, the train will depart the storage lane for the hostler and injected into the mainline when the schedule calls for it. It is a fully automated launch.

The advantages of Option 2 are included but with the added benefit of launching trains in ATC mode from the storage lane; the delay of changing modes from manual to ATC at the hostler is removed, allowing for a faster launch. Training is also streamlined: the train operator only requires training in a single signalling methodology, CBTC. The first two options require training in two methodologies: CBTC and conventional signalling.


To utilize the full functionality of a CBTC solution, the hostler must fall under CBTC control; otherwise the operational advantages of CBTC are lost. The operator’s ability to flag a failed train is compromised, trains cannot be placed in ATC mode until the first station and throughput on the mainline is impacted.

Transit agencies planning to deploy a CBTC solution must understand that a CBTC solution is effective only if it has control over all aspects that affect mainline operations. As a CBTC specialist, it is my opinion that every inch of track should be under CBTC control; a patchwork approach is the main ingredient for operational inefficiency