Hobsonville Point – Why Only 50%

Using The RSP Tool To Settle The Matter

Fast forward to 2024. At issue was the assessment of a 30ha proposed development elsewhere in Auckland. Now having the benefit of qualitative assessment tools like the RSP modeller, someone suggested running it on a 30ha area of Hobsonville Point to provide a real-world benchmark, a comparator that everyone could use to better understand the proposed scheme. Good idea, and not hard to do.

The result? Hobsonville Point scored 44(%). Twice the quality and functionality of a ‘normal’ subdivision (17—23), so, great! But still only a solid… C-. Looking back, it seems like I was pretty much on the button with that intuitive 50%, even if it was mumbled and embarrassingly apologetic.

What’s nice about having the modeller answer the question is that it gives us objective data at two resolutions so we can see exactly what’s going on and where. Looking at the totals, they’re reasonably balanced. Even if you say, ok, the lot and block design could have done with slightly better urban design skills, the real surprise is the open-space scores of 45, 55, 49, 30. These are clearly not acceptable for a top-notch landscape consultancy and for a development with such an outstanding landscape treatment. But then, it’s not the landscape components themselves that are the problem.

Just so you know what the scores could have been, the proposed development we were supposed to be using Hobsonville Point to measure scored 97, 92, 86, 98 for the same sequence. Looking at the field value distribution for Hobsonville Point within each of these criteria confirms the tragedy of the HP scores – that the open-space elements are not bound to their correlating urban form edges. They’re detached from the urban form in a way that diminishes the value of the whole. I’m sure if the designers at the time had something like the RSP tool then not only the faults in regard to the open-space would have been remedied (they involve fairly trivial adjustments with no cost implications), but that most of the urban layout faults would have been fixed as well. There weren’t any missing parts, after all.

What Is It In The Design That’s Wrong?

Would there have been an easy fix? Two design faults dominate.

Problem 1

The first is not understanding nor appreciating the importance of the urban design mantra – fronts to fronts and backs to backs. Put it on a T-shirt, it’s that important. Using mews lanes to strip vehicle movement off the roads is great, but it turns bad when several units are thrown into the mews to find their entries and public interface amid a mass of vehicle manoeuvring and rubbish bins. Adding insult to injury, the mews then explode in size to allow for the pseudo-entries and the space becomes even less habitable. There are techniques to finesse the block shape, block size, and lot plaiting to remove the offence. There are also ‘correct’ habitable solutions for the mews that create life within the space rather than suppress it. Maybe urban design is harder than it looks? Or is it just that the fundamental principles of urban design hadn’t yet migrated to other disciplines?

What lost most value for the designers in the front-back misstep was not understanding that public realm creates the front, whether it’s the area within the road reserve from the carriageway edge to the lot boundary or whether it’s an open-space reserve abutting the lot. This is the margin where social engagement occurs and community is forged. It’s the space that defines and then articulates our relationship to public life.

The biggest failure occurs along the reserve boundaries that are almost all edged by rear or side boundaries. What residents invariably do in this situation is protect their privacy as a priority.

Good planning rules mean they can’t visually exclude the public with high fences, so they plant thick hedges within their boundary instead. This does the job for them, but at the cost of a smaller outdoor living court. Their isolation satisfies a private need, but the effect of everyone doing it is to detach the open-space network from the urban form that should support it. The parts fight with each other. The configuration acts against the concept.

It is not only potential social and community value that suffer in this instance. Public open-space accrues value (and the money to pay for it) in proportion to the number of people who use it. We know that when open-space is fronted ie the front door is accessed directly from the reserve, the household will use the space, on average, eight times per day. Access 30m away (ie one house further back) and usage drops to four times per day. By 200m away usage is considered statistically insignificant at less than once per week.

Although the data visualisation (below) shows considerable dilution of proximity to open-space compared to the proposed scheme, it is at least a tapestry of green with no red lots. The Proximity pattern of the proposed scheme (also below) shows what could be anticipated when an open-space network is embedded as a functional part of the urban form rather than tacked onto it as a bonus extra to be fought against by anti rate-paying resident groups and council park departments with no maintenance budget.

Problem 2

A few lot clusters do correctly front small open-space reserves, although in all cases these reserves are edged by roads along their other sides. This configuration highlights the second major fault – not fully appreciating the extent motor vehicle presence impacts the use of the public realm. The degradation of public realm use in the Proximity measure highlights a spatial relationship; the degradation here occurs as a direct activity conflict.

Over a decade ago the Happy Routes research quantified this effect and showed that people will easily accept an 80% time/distance penalty when a less trafficked route is provided as an option. The same research also confirmed the observation (in NZ) that the oft-repeated 400m ped-shed merrily blows out to around 2km when people have access very high-quality pedestrian-only routes such as the Hobsonville coastal walkway.

The answer to the fault lies in understanding the public realm (of the street) and open-space network are the primary components of the overall movement network. Well, they have been since the national standard was changed in 2010, inverting the mode hierarchy and reflecting what the world had learned about how urban areas work. A movement network is not something that is designed for vehicles that then has a public realm tacked on. The public realm of the street has Place value more than it has a Link function.

While most front boundaries correctly front the street (see the Frontage data visualisation – only a little bit of red), about half of these are also used for vehicle access (a lot of red in the Access visualisation). Medium density narrow lots result in about half of the small remaining public realm area in the street being given up for vehicle function (see the photo below). The street becomes hardly more than a glorified mews and is not an overly pleasant place to spend time.

HLC groomed the builders in their layouts to have an active living room in the house spilling out to a ‘public’ living court at the front (as well as ‘private’ court at the rear). Hurrah! Finally. That’s how you activate the street and do all the wonderful social/community things alluded to above. Unfortunately, when the street has been turned into a vehicle mews, none of this happens. Street culture dies. Those now desolate little front courts speak more loudly of suburban isolation than any massive front set back in traditional subdivisions ever does.

To complete the tale, the proposed development that was under scrutiny (and for which Hobsonville Point was to provide a benchmark) returned an RSP score of 91. So, if Hobsonville Point is twice the quality of a traditional subdivision, it could have been twice as good again – for no extra ‘stuff’ and no extra cost!

Is there a lesson in all this? Maybe urban design is harder than it looks? Or p’raps the fundamental principles of urban design still haven’t trickled down? Maybe. Apparently HP’s known colloquially in the business as the ‘University of Hobsonville Point’. Really? University, with only a C- score? Well, we do the best we can, with what we’ve got at the time. History moves on.

So it goes. (Kurt Vonnegut – Slaughterhouse 5).

References:

• Happy Routes – Quercia et al
• Site Planning – Kevin Lynch
• Life Between Buildings – J Gehl
• A Pattern Language – Christopher Alexander
• Link & Place – Peter Jones
• Neighbourhood-Residents-Questionnaires – Beacon Pathway
• The Role of the Urban Design Review Panel in a Masterplanned Community Development – I-Ting Chuang
• Residential Subdivision Performance Modelling – Common Ground Studio https://www.cgstudio.co.nz/projects/the-laboratory/rsp-tool/