UN 139: Testing Tarneit North FlexiRide - Can it survive a morning peak?

For much of the recent past bus network planning, review and expansion has been out of fashion and relatively unfunded in Melbourne. The government's emphasis was almost entirely on road and rail infrastructure.

Train and bus network reform had a bad name in senior political circles after the 2014 election. Contributing factors may have been Labor's thin electoral margin and a fear that 100 aggrieved passengers would be louder than the 1000 who gained from a change. There were also specifics, for example the backlash arising from Transdev's poorly consulted-on bus network reform attempt in 2015 and Frankston line City Loop fears that set Metro train timetable reform back 6 years. 

Service planning was so much on the outer that there were people saying you didn't need to plan local buses at all. Instead you'd just throw half a dozen minibuses out there with a super-smart app. That can be bought off the shelf with no need to do pesky things like modify roads or install bus stops. The latter would have been music to the ears of a boss who had seen on-ground network delivery capability fall since the Metlink years. It could work as a dynamic self-organising system, with no need for skilled planners to make professional judgements or difficult trade-offs as to route alignments and service levels as would be the case with a fixed route network. 

Fresh from de-regulating commercial passenger vehicles, influential parts of the Department of Transport proved a receptive audience for the tech-bros and their apps who were active spruiking 'flexible transport solutions'. Infrastructure Victoria, which claims to be on the side of economic efficiency, productivity and social utility, has also been favourably disposed towards flexible route buses. 

Early claims were made of FlexiRide's success in Melbourne's outer east. However these comparisons were against the low productivity flexible route TeleBus services they replaced rather than the generally more productive conventional fixed routes. The latter would have been a useful test, especially when deciding between the two approaches in large, growing and dense neighbouhoods. 

The government took up the demand responsive thing with gusto, introducing FlexiRide buses to more suburbs across Melbourne, despite potential reliability risks and opportunity costs. The suburbs that got them varied considerably, from older and sparser Rowville and Lilydale to young, dense and diverse Melton and Tarneit North.  

During all this too few in authority asked 'Do flexible route buses work?' And, if so, 'Under what conditions?' and 'How well compared to alternatives'?

I will answer these questions today. Firstly, I can vouch from experience that FlexiRides are wonderful - if you are the only passenger. 

But add more passengers and it degrades before falling over. Firstly in having to deviate via other peoples houses, extending travel times and risking missed connections. If that happens too much then you are better off with fixed, reasonably direct routes. Secondly the waits can get so long that you might as well have a fixed timetable so you at least know when the bus will come and it can be optimised to connect with trains. 

You can overcome both if you add more buses and drivers but that  increases costs. And, unlike fixed route buses, which get more cost-effective as more people use them, this isn't so much the case with flexible routes which are not so scalable.   

 

Tarneit North tests

It's all very well to mention these limitations but where's the evidence?

Last month on one weekday morning I used the app to test the performance of Tarneit North FlexiRide (which started just recently in October 2022). I did this by planning hypothetical trips mostly to and from Tarneit Station.

I started this test early. Wait times were often around 10 minutes. Not bad. However they tended to increase as it got later and more people wanted to use the service. Below is a counter-peak direction trip from Tarneit Station. It had an advised pick-up time of 16 minutes.

The trip below is an inbound trip somewhat later (around 9am). The estimated wait time was 35 minutes. Travel time to Tarneit Station was estimated at 17 minutes. Thus it would take the better part of an hour to go a few kilometres to the nearest station. 

What happens if all the FlexiRide buses are used and it's not considered worthwhile to deviate one of the five buses out? Now you get a 'no available vehicles' message. The app include fixed routes with a suggestion to take Route 182 departing in 19 minutes. That doesn't go very near the destination with a 39 minute walk the other end. Alternatively, and quicker, you can just give up on the bus and walk the entire 3.7 km to your destination. 

Both the last two trip plans were around 9am, which appears to be about peak demand for FlexiRide. As this subsided it was again possible to plan trips with shorter waiting times. More discussion here: https://twitter.com/MelbOnTransit/status/1595159838782545922

Conclusion

These tests demonstrate the basic volatility and unreliability of FlexiRide-type services in populated areas. Having the service come up as being unavailable at certain times indicates that it cannot reliably survive a morning peak in areas with significant travel activity like Tarneit. 

Thus FlexiRide does not scale up well. That's a big problem if you wish to efficiently introduce buses to a large and growing area where you need to be able to satisfactorily serve the most numbers of passengers for the least cost.  

It's true that existing fixed routes also degrade in performance when a lot of people use them. However this degradation is more gradual and starts at a higher patronage level. Unlike FlexiRide, which degrades at an earlier point, regular fixed route buses represent efficient passenger conveyance over a wide range of loadings, from several to several dozen. This wide versatility is one of the under-acknowledged benefits of fixed route buses. 

Tarneit is the sort of area where you can put on a conventional route bus, even if not very frequent, and people will use it at rates higher than in many other parts of Melbourne. This involves passenger numbers more suitable for fixed routes. Thus, instead of FlexiRides, I would favour fixed routes introduced as early as possible and run as frequently as possible. 

Flexible routes may have a use where an estate is very new and only a few scattered homes are inhabited. But as the road grid is completed and people move in then they are past their use by date. If flexible routes are introduced at all I suggest there is an exit strategy including close performance monitoring and a trigger where they get converted to fixed routes before availability and reliability degrades. For this reason performance data should be published for FlexiRide as it is now for fixed  bus routes as public accountability is currently lacking. 

Infrastructure Victoria correctly acknowledges in their recent bus discussion paper that flexible route buses are most appropriate for low demand or niche applications. However given actual performance I still think they over-sell their applicability, especially relative to fixed routes which can more reliably carry more passenger per hour. This is something you might wish to comment on if you put in a submission (deadline January 27, 2023). 

See other Building Melbourne's Useful Network items here