Route 90 snapshot: 24 March-20 April

This article presents a "snapshot" of service quality on Route 90 between the 24th of March and 20th of April. Data sourced from Transsee. 

Cancellations

During the analysis period, the overall cancellation rate was 5.4%. Northbound trips saw a slightly higher rate at 5.5%; southbound trips sat at 5.2%. 

Of note - the 27th of March is without data, the route data that day is extremely poor. The 8th of April saw a big snowstorm, with a crash on the Kichi Zibi Mikan causing cancellations across the bus system. 

Headways

Week 1 overall saw poor data quality (the cancellation figures are hand-checked). The other weeks have a solid quality of data. 18 April is excluded from Week 4 in all of these charts as it ran a Sunday schedule for Good Friday. 

Northbound first. 

Excluding the bad data of Week 1, we can see a clear pattern of bunching in the AM peak growing over the route's length. Additionally, bunching increases further north along the route, most acutely during the PM Peak. 

As we'll see later, Route 90's schedule should be much improved with New Ways to Bus coming soon (as of writing), but the current schedules are poorly written, despite low traffic levels along the route. 

Southbound trips show a stronger bunching pattern, one which also grows over the length of the route. Bunching is already evident at Billings Bridge, but it becomes stronger towards South Keys. There is evidence of bunching throughout the day, but like most routes, it is the most evident during the peak period. 

On-Time Performance

A quick look at the performance charts suggests a striking pattern. 

On-time performance is good at the origin, but it drops quite dramatically, especially NB but also SB (SB has some recovery towards the terminus). 

What these charts (and more to come) will show is, the current schedule for the 90 is poorly written. New Ways to Bus should deliver genuine improvements to Route 90, but if the interlining soup continues, and I will analyze that at some point, then delays from other routes will continue to propagate onto Route 90. 

Already, on-time performance is poor in the AM Peak. This has repercussions both for the route itself but also for the routes interlined with Route 90, as their trips might be affected as well. 

Northbound, the "peak" direction, is particularly bad, with the trips consistently falling behind schedule along the entire route. Southbound trips are only relatively better in this respect. 

Midday trips show the same lack of scheduled runtime, though I have not analyzed them the way I analyze peak runtimes. Once again, we see an especially poorly written schedule leading to cascading delays along the route, which steadily grow over the length of the route. I suspect the poorer performance of SB trips leaving Hurdman is due to late NB trips arriving at Hurdman, possibly with inadequate layover time as well. 

The PM Peak shows the same terrible performance as the rest of them, again with NB trips (this time the counterpeak) with worse performance. Extra time is scheduled SB on Downpatrick and Hunt Club. 

We can also look at the string diagrams. I have selected week two to pull. The weekday PDF can be found here and the weekend PDF here. 

Here's a look at Wednesday: 

There are fewer pinch points on this route than other routes. Here's a few, though note that the magnitude of these is smaller than many other routes and could be improved by tinkering with light cycles: 

• the Hunt Club/Uplands intersection 
• Turning left from the Transitway onto Hunt Club (Update: it was pointed out to me that this intersection does not feature a left turn - I had forgotten about the SB->WB ramp. There seems to be a delay here going NB, which may be the Airport Parkway/Hunt Club intersection.)

Surprisingly, Heron does not seem to be a chokepoint. 

This route should be reliable, and you will see why. 

Travel time

Overall, you may note the anomalies in the travel time; one is a definite error, and others may be as well ... this is an area of future consideration for myself. 

Otherwise, travel time is generally consistent across the route. It's good to see improvement in the schedule, but this is something that needs to be regularly and consistently done across the network, not just with major service changes. 

This is the NB direction in the AM Peak. The current schedule is totally inadequate to runtimes, resulting in the on-time performance graphs you saw above. However, the travel time is still inadequate south of Uplands/Riverside after New Ways to Bus - the refusal of OC Transpo to run adequate runtimes in the peak direction is one of the more perplexing decisions of New Ways to Bus. 

South of Uplands/Riverside, the current schedule was not once met over the four weeks, while it remains firmly in the bottom quartile north to Billings Bridge. No wonder we have a reliability problem. 

Southbound, the AM travel time is somewhat better than in the NB direction. The travel time has clearly been improved where it's needed, which is good to see - this should reduce propagating delays on Route 90. 

Looking at the SB trips in the PM peak, the travel time is again inadequate north of Uplands/Riverside and well-scheduled - perhaps excessively scheduled; but OC Transpo is prone to underscheduling elsewhere, so it can fly. 

Once again, good improvements to the schedule with New Ways to Bus, which will hopefully reduce delays on this route. 

Let's look at NB ... 

The current schedule is completely inadequate for northbound travel, which reflects what is seen in the OTP charts. This route could benefit, and hopefully, will benefit from an increase in scheduled travel time. 

 

Thoughts

Route 90 is a prime example of how scheduling can make or break a route's reliability. The bunching is not severe outside of peak periods, but there is a low level of bunching "noise" at all times; however, the routes run behind schedule very consistently, despite low levels of congestion as indicated on the travel time charts. 

New Ways to Bus should deliver genuine reliability improvements here (as opposed to mirages such as Route 85). This is an exercise that should be repeated across the entire network, even at the cost of service - though I believe there is a way to improve service at no cost ... more on that soon (hopefully!). 

Like Route 51, this shows the importance of keeping a good schedule in order, as delays can propagate, and even with no traffic, the favourite blame of OC Transpo and advocates alike (a great work of deflection, I would call it), service is still unreliable. 

That's it for Route 90. A short piece on New Ways to Bus thoughts coming soon, and then hopefully, a longer one on OC Transpo scheduling practices. 

Until next time.