Urban Filler is a computational urbanism experiment, which takes the viewer a long way from the ideal of sociable city via the possibility of urban densification to the discussion of the current regulatory regime. As urban planners will attest, infill development is challenging: the local residents resist and regulations restrict. We explore the constraints that regulations create with the example of a single urban block in Myllypuro.
Introduction to infill development
Urban life in Helsinki comes in many forms. The centre naturally offers the most and possesses the historic urban fabric; the distant suburban areas provide the possibility of the living in one’s own home; and in between them lies a ring of the more recent history, large housing estates of 1960s -1970s. Appealing as they were to newcomers from across the country during rapid urban expansion of that period, they present a challenging case today. They neither possess the urban life we find in the central, traditional city districts, nor do they offer the attraction of detached housing. Instead they are a city environment, without the city life. They are typically a big open structures with the floor-to-area ratios of only 0.5-0.6, which compares unfavourably with a much lower-scale and closer-knit historic districts, such as Vallila. Their density is also lower than in the present-day mixed residential areas, such as Arabianranta, where the floor-to-area ratio reaches 1.5 (Figure 2).
A possible response to the challenge is revitalisation of 1960s – 1970s suburbs through densification. It will be logical to complement their structure with infill development, turning them in the process into decidedly urban environments, which could sustain street life in the same way as traditional city districts do.
Remaking urban environment, however, is difficult. The infill projects are more demanding technically, and most crucially, they depend on the support of local residents and condominium companies who own the existing real estate lots. The process, therefore, has to begin with creating interest among stakeholders. Urban Filler takes on this task by communicating the potential urban scenarios to the stakeholders. It visualises the volume and configuration of infill buildings for a different levels of height restriction and parking provision. The demonstration of relationship between the regulations and the resulting building volume will help the stakeholders to understand the trade-offs between provision of parking and overall density, and explore the effects of current regulation on infill projects. The visualization is implemented in the form of interactive display where the viewer can choose the level of building height and parking space and see the resulting infill building volumes (Figure 1).
In the project we produce infill scenarios for an urban block in the central area of Myllypuro district. The district is one in the series housing estates stretching along the metro line in East Helsinki. Its core was built in 1960 and it consists of large multi-storey apartment blocks set in an open plan in a pine forest. Although beautiful in some way, the district could benefit from additional activity and services, which intensification projects will bring. Presently Helsinki City implements a series of projects in Myllypuro. In place of the old shopping centre next to the metro station, a large complex of housing and service centre has been recently finished and there is an area of detached housing under construction in the south section of the district. There is now a new fabric both at the very centre and fringes, however most plots of multi-storey apartment blocks estate in between remain in the original form as designed in 1962 plan for the area (Figure 3). Continuing the densification from the metro station, we selected the next block to the west for our study.
We decided to investigate using the ESRI CityEngine generative modeling package as the platform for producing the scenarios. The software uses CGA-scripts to generate building shapes according to specified rules. Our initial goal was to produce a script that would identify suitable locations for infill buildings within the existing plots, but this proved impossible due to limitations in the way CGA-scripts operate in the CityEngine software. We resorted to specifying the location of the infill buildings manually and concentrated on exploring the effects of two urban planning regulations on the urban structure.
Block architecture & generation algorithm
As the urban planning rules to explore, we selected parking regulation and maximum number of floors. From our discussions with planners at Helsinki urban planning office, we learned that the requirement for a minimum number of parking spaces is a key constraint in infill projects. According to the regulation, a certain number of parking spaces must be built based on the gross floor area and/or number of apartments (link to reg). Maximum number of floors on the other hand is a typical urban planning constraint and it has a significant effect on the architectural character of the block.
In the scenarios new infill buildings are located on the edges of the existing superblock. In this scheme views and access to sunlight have been preserved for the majority of current residents. The new buildings complete the sparse forest suburb block into a sheltered and more urban configuration.
The scenario generation algorithm takes the user supplied values for the regulations and “inflates” the buildings to produce the maximum gross floor area allowed. According to our calculations a maximum of 155 parking spaces can be accommodated on the plot without sacrificing too much area. This number serves as the constraint on the side of parking regulation. These spaces must suffice for both the existing residents and the new ones.
Discussion & future work
The scenario modeling shows that the current applicable parking regulation of one parking space per 120 m² of gross floor area basically prohibits any densification even on spacious plots of forest suburbs. The ~1600 m² allowed by the current regulation would amount only to three 2-floor buildings or one 6-floor building making the plan economically unviable. A less strict regulation of one space per 140 m² would already allow ~5000 m², providing homes for ~100 new residents.
Due to the fact that we were unable to generate scenarios for the general case, this study is limited in scope. More advanced methods such as constraint solvers or evolutionary algorithms should be investigated to approach the problem of identifying suitable locations for infill buildings.