At the TTC Board meeting of January 21, there was a presentation based on the report Safe Service Action Plan that included several actions and proposals that are intended to improve the safety of TTC operations. The context for this report is that there have been a number of high profile incidents involving TTC vehicles in the recent past.
During 2014, two pedestrians died after being struck by streetcars. Then, in the latter half of 2014 several video recordings were made public of TTC bus operators running red lights. In response to these incidents, the CEO initiated a review of operator training, supervision and relicensing as well as a communications campaign to reinforce the need for operators to drive defensively and to adhere to the rules of the Highway Traffic Act.
Towards the end of 2014, an adult woman and a 14 year old girl died as a result of injuries in separate incidents after being struck by a TTC bus making a turn. Given the very serious nature of these tragic events, the CEO directed that the review already under way be expedited and that it include consultation with other operators for comparison and to seek out best practice. [pp 1-2]
Broadly speaking, the proposals can be grouped into three sets:
- Operator training and recertification, including an emphasis on safe driving.
- Monitoring systems to track and record road conditions with a view both to increase evidence when accidents occur, and potentially to monitor driver behaviour.
- System review including issues such as stop placement, identification of accident “hot spots” and causes, and improved investigative techniques.
Two things were quite notable in the presentation.
First off, there was almost no reference to consulting with drivers as part of the system review. They would be chatted to about the importance of safe operation, retrained regularly, and monitored, but the feeling was of a very top-down process. A member of the Commission, Anju Virmani, raised this issue, and CEO Andy Byford replied that management has a very good working relationship with the operating staff. This position was not shared by ATU Local 113 President Bob Kinnear who argues that driver behaviour is influence by heavy-handed service management. Byford claims that his message is that “safety takes priority over the schedule”, but from the union’s point of view, that’s not what actually happens.
A long-standing problem has been the suitability of scheduled running times for actual conditions on routes. The TTC has had the ability to monitor and analyze vehicle tracking data for years — I started doing this in 2007 and the system whose data I used was hardly new at the time. Some routes and time periods have lots of room in the schedule for their trips, others are not so good with the result that operators are pressed to make unrealistic schedules. With little headroom under normal conditions, any disruptions inevitably bring on short turns to keep the operators on time.
In the fall of 2014, running times were increased on some routes, but this was done more-or-less across the board during all operating periods. The result is that a route like King now has cars dawdling along the route and queuing up at terminals. The flip side, in theory, is that short turns are much less commonly needed for cars to stay on time, but the speed of service has declined. From a rider’s point of view, it is hard to say how this balances out, but giving vehicles far too much running time is simply a case of cooking the schedules to produce a desired result (fewer short turns) without considering the balance between reasonably speedy service and adequacy of terminal breaks for drivers.
The logical conclusion would be that the TTC should actively work on revising its schedules, a process that is now underway. However, in a bizarre response to the claim that operators drive too fast in order to stay on time, the TTC will experiment with covering up the dashboard display that tells a driver whether they are late or early, and by how much. That’s right. The very mechanism that the TTC might use to manage service and ensure reasonable vehicle spacing will be disabled. A trial of this will be conducted on the East Mall bus although, with its double-digit headways, I am not convinced that this is a representative route.
As if that were not strange enough, there was a reference in the presentation (not in the printed report) to a proposal that streetcars cross intersections no faster than 25km/h. This sounds sort of reasonable if it were to apply only at major intersections, especially with track junctions where there is already a go slow operating standard. But no. All intersections would be treated this way, even the most minor, and this would effectively mean that streetcars would never run faster than 25km/h.
What this shows quite clearly is that whoever dreamed up this idea has never been on a real streetcar route in his life. The average speeds may be in the low teens, but the top speed regularly gets above 40km/h especially away from downtown and in off-peak periods.
If such a proposal were made for bus routes (where the real problems with speeding and running red lights have been reported), it would be laughed out of the room. The Lawrence East bus, for example, routinely tops 50km/h.
There are times that the TTC’s attitude to “safety” is selective with the word being used as a motherhood shield to limit debate, or to strong-arm acceptance of badly thought-out schemes such as the interpretation of fire codes in the design and placement of second exits from subway stations.
Oddly enough, a long-standing issue on the streetcar system has been ignored for a few decades — the operation of electric track switches. When the ALRVs (two-section streetcars) were introduced, the old track switching system had to be replaced because it depended on all cars having a common distance from the front of the car to the point where the trolley pole touched the overhead power wire (the switches were actuated via a circuit on the trolley pole connecting with a contactor on the overhead).
The then-new system depended on in-pavement detection loops and radio antennae on the streetcars. This system did not work well when it was installed 30 years ago, and it has become less and less reliable ever since. All switches on the system, even those that are manually operated, are now treated as a stop-and-proceed situation rather than a slow, rolling approach, with the result that operations through intersections can be jerky and tedious. The south side of Roncesvalles Carhouse, westbound, is a particularly bad location.
There has been a project on the TTC’s books for some years to replace the existing system, but this has been deferred many times. Implementation of a replacement system may occur over 2015-16 assuming it is not pushed off again. Aside from the question of reliability and the fact that many nominally electrified switches are now “temporarily” out of service, this increases the number of locations where operators must alight from their cars, throw switches (in whatever weather and traffic conditions), make their turn, and walk back to reset the switch to ensure that the next car doesn’t come upon an open switch. On top of this, locations where the track switch electronics might interact with traffic signals to give a transit-only turn phase don’t work if the switch is manually operated. Diversions, a common enough event, are the very time when automated switching and transit priority would aid an already difficult situation, but these are often hampered by being fully manual operations.
Another source of delay is the lack of priority signalling at many locations where it once operated. The whole idea, many years ago, was to reduce running times, and in particular to avoid delays to streetcars (and later buses) caused by the fact that they move more slowly than regular traffic because they must stop for passengers. As the number of working intersections declines, running times go up.
I have no problem with the TTC pursuing a goal of safe operation. They should be a model for all other motorists, and should protect both their riders and their staff to the greatest degree possible. That said, the “safety” program can be compromised by
- appearing to selectively target staff monitoring rather than consultation,
- treating service reliability as a secondary issue to the point where it can be ignored, and
- failing to address underlying problems such as outmoded schedules and a traffic environment that works against the best possible transit operation.
It is no “improvement” to hobble service with unreasonably low speeds and by throwing schedules out of the window. They are both part of the service a transit system provides.
A safety mantra is not enough — it must be backed up by a culture that views blame and discipline as an exception for the very few “bad apples”, with collaborative improvement on and by all sides as the goal.
Appendix: What are typical operating speeds?
The charts (an Excel spreadsheet) linked below give details of average operating speeds on 504 King for one-week periods. Note that this chart has multiple tabs, one for each hour of operation.
They are created by the following process:
- Vehicle tracking data, including GPS location, are transmitted every 20 seconds.
- These locations are resolved to a scale on the route with one unit per 10 metres. This is a choice in my analysis, not a limitation of the data.
- The change in position for each “tick” of the clock is a measure of the vehicle speed for that interval.
- Values observed at each 10 metre increment are averaged within each hour to produce the data that populate the charts. (Some points will have no data, especially during periods of wide headways, because no vehicle reported being at that spot during the hour in question. A vehicle travelling at 40km/h will travel 222m in 20 seconds, or 22 units. It will, therefore, be “seen” at under 5% of the 10-metre segments it is covering based on a 20 second reporting interval.)
What is visible are the locations where operation is slow, particularly where this occurs over an extended distance due to congestion. It is quite easy to pick out the location of well-used stops because there is a notch showing a low average speed because most vehicles stop there. Stepping through the tabs gives a flip-chart animation of changing conditions throughout the day.
It is quite easy to see that 25km/h is too low a speed for the King car in many locations and time periods.