I am going to start my presentation, as one often does, about a personal experience that opened my eyes a little. I was invited to talk about my design process – which I found interesting, as the audience would be CEO, CFOs, business managers etc. – all undertaking an executive MBA course. What was fascinating about all this of course, was that what we (designers) have been doing for years, often with little self-introspection or awareness, was in fact revolutionary in some circles.
Indeed, writing in the Harvard review in 2018, Jeanne Liedtka referred to design thinking as a new way of organising work, that leads to extraordinary improvements. Whilst I won’t go into much detail here, typical analytical thinking tends to identify an issue, throw up several solutions, discuss each on its merits, and then pick whichever logically gives the best outcomes. The issue is, solutions tend to be considered in isolation, and therefore, innovative ideas that are previously untested tend to be eliminated quickly.
Design thinking looks at the problem differently – it actively seeks to design an outcome that draws on many data points, combining and revisiting ideas as they are evaluated and tested, to create a new solution rather than choosing one from a list – a little like the difference between buying bespoke furniture or something ‘off the shelf’. It is a non-linear process, and invites creativity (unsurprisingly, given the name). Interestingly, for something I have never codified in my own work or that of my colleagues, many articles refer to the key stages as being able to empathise, define, ideate, prototype and test. I think that first word, ‘empathise’, is really at the core of what we are looking at, when I do finally get around to the core subject of today, which is of course, Open-source Urbanism.
Design thinking has now well and truly ‘crossed over’ into the mainstream, so let’s look at the flipside.
The term “OPEN-SOURCE” comes from the software industry – now I’m not an IT expert, but an architect and urban designer – so please bear with me on this. Open-source is the natural growth of ‘freeware’, which has been around since the 1980’s; however, with the growth of the internet in the 90’s, and the ability to collaborate outside a singular workplace environment, we started to see the emergence of something new… open-source software.
What sets ‘open-source’ apart from mere ‘freeware’, is the democratisation of the creation process. Freeware was generally as it sounds – software written (by either a sole programmer and/or an associated team of programmers), then freely distributed – ‘open-source’ started literally opening the source code – allowing programmers from anywhere to access the code and debug, improve, develop, or even completely transform the usability of the software. Certain methodologies or protocols are often used to ensure that someone doesn’t just ‘go in and muck it up’, but of course, as the code is iterative and open – one can simply go back to earlier versions, or only use the source code they require. In most cases, it is copyright and royalty free… as the idea is to get the best ideas, no matter where they come from, and build upon them.
This allowed ANYONE to contribute; quite organically, the best ideas were picked up, re-used, and incorporated, whilst buggy or poor script was generally overlooked, and lost or taken out.
Out of this movement, we gained several pieces of software… even operating systems that we know today; Linux, of course, is possibly the most famous, whilst other names such as Mozilla (once more popular than Microsoft’s own Internet Explorer – 2009), Word Press, Apache Open office, Ubuntu etc., are all generally well known, and seen as real drivers of innovation within the tech industry etc.
This ability to draw together interested parties, stakeholders if you will, and develop the software without an overriding corporate agenda, is something that set an example (as did design thinking) for other industries… and we, as designers, started to gain inspiration.
Moreover, we started to re-assess not only our methodologies, but where we gained our raw data, our ‘source code’ if you like.
So why am I revisiting software before I get onto urbanism? Because there are some great lessons we can learn here… and I believe it gives us some insights into how open-source urbanism can develop in the future… for the right reasons – and not just ‘because we can’.
Open-source software opened the creation process to all – what I refer to as the ‘democratisation of the development process’. This allowed good ideas, innovation, development, even simple ‘cleaning up’ and efficiency gains, to come from anyone with an interest and the skills to do so – not just the employees of a single entity or team.
Opening the process removed institutional bias: because people were working out of interest, often unpaid, they weren’t serving multiple masters – they were effectively only looking for the best solutions, not necessarily the most marketable, or those that fit a pre-determined vision.
Ideas were able to be tested by peers, almost in real-time, ensuring constant refinement: as multiple programmers were sharing code, they would take lines of code and immediately test them or use them, then incorporate, reject, or improve on any code… or even re-use and repurpose certain code for other functions.
Within our industry, which I will loosely describe as the Built Environment and Place Making – arguably (and someone will always argue on this) …the first to really jump in… was architecture.
There are some great examples that have come to prominence lately, especially with architects providing free plans and details, particularly for things like low cost and social housing, flood resilience, and emergency housing etc.
Work like that of Pritzker prize winner, Alejandro Aravena, who released the plans of Quinta Monroy Housing in Chile… an incremental project that grew over time; thus, like open-source IT, designs were able to be taken, used freely, and improved or developed over time – importantly, Aravena made the high-quality architectural designs accessible via downloadable AutoCAD files, now part of the public domain via creative commons licenses.
Of course, in some ways, open-source has been available in that form for a great many years – one needs look no further than Christopher Alexander’s seminal work, “A Pattern Language”. This of course, gave solid design examples of how to approach a design problem, typical patterns of solution, and then some research to back up the claims, so the ‘lay-person’ could easily understand and follow; no need for a design degree to understand the book – it’s almost a self-help design guide to anything from a room to a house… indeed, right up to neighbourhoods and cities.
This kind of open-source is very interesting to me, but what I find differentiates it from the open-source we have seen in coding, is who is writing the code. I applaud Aravena and Alexander… their work, and that of others, is genuinely interested in making architecture not just available to the rich, the developer, or some kind of government authority, but ultimately, these well known, well-educated design professionals, are simply making their research more available in the general sense… codifying their learnt knowledge (largely via an institution and through practice), and formatting into something that can be repeated and adopted by non-professionals.
This is repeated in other open-source projects. A quick google of examples of open-source architecture returns several results – such as Open Desk… who give out CNC router cutting files, so anyone can make their cut plywood furniture, Aker do a similar thing for gardening equipment, whilst Paperhouses, like Aravena, make plans available through PDF, and even DXF files.
To me, this is extremely laudable, and vastly improves access to quality design. Indeed, it keeps professionals like myself on my toes… as we always have to produce better and justify our fees. However, what is somewhat (but not entirely) missing, is the contributions at the creation stage… the “INPUT” stage… the source code if you like – one could argue that this is open-ended architecture, rather than open-source, because that ‘source code’ is very much coming from companies, professionals, and theorists; although Alexander would argue his ideas came from repeated patterns, and the public were the judge of what works – hence, the source code is not entirely missing.
One example that I believe is true open-source architecture, was never presented to me as such – it was more an issue of understanding the vernacular. At the Australian Architects conference many years ago, founder of Studio Mumbai, Bijoy Jain, recounted a story that has stuck with me, and influenced me ever since.
Should Jain hear this, or if this should get back to him (unlikely), I hope he forgives my recollection here, as I’ve only ever heard this story verbally. However, to my best recollection, he explained that when using bamboo staves for a fence, his team would only cut the staves during a full moon. Being a man of education, he laughed off the superstition… as he saw it, and the fences rotted. But when he allowed the team to cut them on the full moon as they wished, and according to tradition, the fences lasted, and did not rot. He later did more research and found out that a certain insect that lived in the bamboo would vacate at full moon – presumably to move to new ‘lodgings’ – so that bamboo cut at that time was ostensibly parasite free, it was clean quality bamboo – by sake of a natural cycle. Whilst the science of ‘why’ was possibly unknown to the local team – the practice and reality certainly was.
By seeking the ‘source’… the actual methods of production and build, Jain was able to benefit from knowledge he did not formerly have and was able to incorporate that knowledge to produce better outcomes.
The ‘source’ was not from a professional body, a research team, or through professional designers, but the benefit was real. Jain then went on to build more complex structures and push the material into more complex and inspirational forms and uses… because he could rely on the material – something he had not done before, as it was considered unreliable and subject to rot. That one bit of ‘source code’ had unforeseen benefits, and was able to be built on, expanded, and to provide real benefit – because the architect, the professional listening, had ‘opened’ to knowledge from an unlikely quarter.
This finally brings us to Open-source URABNISM!
Before I dive head long into the logistics… what is being done today, what is developing etc., let’s look at why I have taken you through this re-cap.
My big take away is that the difference between open-source architecture and open-source code is the input level… the ‘source’ itself, if you like. Whilst it is great to share our code – it is also, I believe, imperative to get collaboration and input at the creation stage, and during the development of concept stage… not just to share our work. I would argue that the greater the input from invested parties… and from unlikely sources, the greater the impetus towards innovation, and most importantly, the response to real need.
That input does not need to come from professionals alone (don’t worry… you’re not out of a job… more on that later), but from inhabitants, users, and those most affected – particularly in the case of urban design.
One of the largest challenges we have as urban designers is understanding and comprehending the needs of those who will occupy, use and develop the spaces we design.
By cracking open the source, listening to client direction and really democratising those initial inputs, we open the feedback loop to those who matter the most… the end user.
Let’s look at Open-source Urbanism as it stands today.
The first thing I have to say, is I do believe it is open-source, not just open output – in fact, done right, it is a feedback loop that takes constantly evolving inputs which produce outcomes that then become inputs for the next generation of design or development… even of the same urban space – which is EXACTLY why open-source programming and code has been so successful.
However, I must interrupt myself, and say that open-source outputs (sharing knowledge) are fantastic – just as it has shown to be beneficial in Open-source Architecture.
Sharing knowledge, technical specifications and solutions is a wonderful thing; the collective knowledge of the urban design community can be disseminated and used by a wider audience, ultimately leading to better urban spaces – it’s not however, a one size fits all approach.
Let’s take a simple example – an aspect of the city that has been desired by progressive development authorities for years now is Biophilic design. Here we are in Singapore… where Biophilic design has been championed. We see it in buildings, in playgrounds, and in public spaces… it’s brilliant! As a designer working mainly in Viet Nam and Australia, there is a lot of inspiration to draw on. However, if we simply just applied the hard details, the actual solutions may work in Viet Nam, but they are likely to fail in Australia, simply due to climatic differences and eco system variability. I would even argue, given the amazing landscape team I work with, that whilst similar in many ways, changes would also need to be made in other Southeast Asian countries, compared to Singapore, in order to achieve the outstanding results we see here.
This then, needs to be treated as an input. However, it is already coming from the larger design industry rather than the community itself, the ones who will ultimately inhabit the space. But… we do have methods of getting that open-source community input – that source code if you will.
So, what are those inputs?
Traditional sources like community consultation is 100% open-source. Organic community input is open-source; for example, a group of local activists lobbied for a community garden in Perth when I was still there. They enlisted my help to draw some plans… it was all very grass roots and low key… that was 15 years ago. Since then, it was built as a trial, the land protected by the authority, and is now an exemplar, that has been copied and replicated, and incorporated into urban developments all over Western Australia. The urban design community… the source code if you like, was simply the community itself, wanting a space to garden together.
Other inputs are things like:
A lot of these have become somewhat digitised, to allow for access to the most current and recent feedback in a timely manner – even opportunities to interact and respond to community concerns, wants, and needs.
For example, Wheelmap is a program designed to find and assess wheelchair accessible places around the world. Urban designers can access this platform and not only see where improvements can be made when rejuvenating built areas, but also use it as a tool to identify what works and analyse that for future development.
OpenStreetMap is another tool that allows community input and as such, identifies trails and walking paths, and non- commissioned roads etc., that may not otherwise be found through more official sources of information. It can give real insight into circulation and traditional paths that may not otherwise be known to the urban designer – it opens the source code to the inhabitants, so others can gain greater understanding.
But there are limitations with this traditional approach. There is a tendency for self-interest. In some cases, groups form to hijack or ‘game the system’.
There is a tendency towards ‘more of the same’, and of nostalgia, which can stifle innovation or big picture thinking.
Moreover, level of detail is difficult to manage. For example, one person may understand traffic flow in their street and neighbourhood but not the interconnectivity to a larger picture or regional transport system… yet that voice can get very loud and drown out the needs of the larger community.
Even in an open-source model, without careful management, those same voices with relevant concerns can be as easily lost – just like a more traditional linear model, where the research is top down rather than bottom up.
This, however, is just a one part of open-source urbanism – it is a modernisation of traditional community consultation, and it is slowly and surely improving and opening access. However, like a letter, or even an email, where one communicates directly, it is somewhat stuck with its roots in the past. People are learning and giving feedback every day, often in unknown ways.
Data mining on social media (and various other online platforms) tells us what people are looking at, what they are buying, where they go, what routes they take to get there – we all know this. The difference is, this does not require active participation, and the data pool gained by these methods are vastly larger than a survey of how people act for example, which not only relies on people’s recollection, but also their desire to engage – I mean who has time to sit down and fill out a survey?
The next stage of open-source urbanism is to take direct data from the urban space itself – to constantly update and refresh the information presented and use that to make informed decisions, based on the actual (and very real) actions of the inhabitants themselves.
This can and is being done in a plethora of ways, from the aforementioned data mining of online platforms to transport patterns generated daily via Google maps, and even Facebook and TikTok geo tags that not only track locations, but movements and behaviours.
What we are also seeing embedded into smart cities are open-source common urban assets, that are co-produced by both professionals (in many fields) and inhabitants themselves. What I mean here is that data is constantly being collected – by stores, by businesses, by ticket machines, by toll booths etc…. even by the toilets here at Singapore airport (did you enjoy your experience?). On top of this we have water usage patterns. This information gives a wholly unbiased, real record of behaviours and interactions. It is constantly evolving in real time and when extrapolated over months and years, can be analysed for seasonal effects or changing habits.
Furthermore, interpolating data can give insights as to why and how people use spaces. For example, drawing on many data sources may identify that people like to walk a certain path, but slowly, indicating leisure time, and therefore, this may be a place to provide services or a kiosk… as in a different data set, it may emerge that drinking fountains in this area are most used, due to an absence of food and beverage services.
I have been fortunate enough to work with colleagues on smart city initiatives, particularly in Northern Viet Nam where we were planning exactly that… a smart city. Working with Aurecon, this embedding of feedback loops into the urban fabric itself, gave an opportunity for unprecedented feedback on the urban design and systems to be realised, and the opportunity for fine tuning and tweaking as the township developed.
Some of the embedded initiatives were:
Energy and water use metering
Monitoring and analytics
Other urban artefacts can also be used as data collection systems; for example:
Lockers at shopping centres
Kinetic paving tiles
There is another approach to open-source urbanism which is, in a way, to provide the source code in the form of built infrastructure, or partially built infrastructure, that can be adapted and adopted over time by the community.
These things may be landscapes and public spaces, that could be used in multiple ways (Hassel’s work in Shenzhen comes to mind) …effectively designing spaces deliberately to allow (to use Bruce Sterling of Wired’s words) “creators to build on each other’s work… co-opt, repurpose, and remix in a decentralized way, creating original products”, or in this case… original outcomes.
Other items included into the built structure can be buildings that have no fixed purpose. In my former practice, we looked to repurpose a large cement factory known as the “gantry”, that overhung the Saigon River. The truly interesting thing was that the developer only wanted to spark ideas for use and leave it to the market… to the people, to determine its amenity. The idea was an expansion of co-working to co-habiting, co-entertaining – really co-living in all living’s facets, and it was the core of the masterplan for the community. To leave such things to the community requires vision and faith (unfortunately, in this case the previous owner demolished the building, considering it an eyesore and not worth saving prior to handover of the site!).
This is, of course, all fantastic. We now have a huge feedback loop of data. We have multiple methods of input from active systems such as Wheelmap or Community consultation surveys, to passive systems that collect digital data, or direct behavioural or physical influence on the built form… in addition to data mining and analytics through digital platforms already in use.
Apart from issues of privacy (which is another presentation in and of itself), the issue becomes one of management, because whilst I’ve talked about what open-source urbanism really is, the truth is…
…that to be truly open-source, we open ourselves up as designers to what may be an insurmountable volume of information.
Obviously, we need to streamline how we deal with the data – some is easier than others. With built interventions, such as landscape spines or open use urban elements, the outcomes are based on the community, and not prescribed by the designers. These are more like providing the source code (as we see in open-source architecture), allowing the community to take ownership and to develop them. This is a more set and forget approach, combined with an ‘observe and learn’ mindset.
However, with the community consultation, with the data collection, and the smart initiatives (I’m sure we’ll hear a lot about this week) …that is the source code for us as urban designers to draw upon. The question is how and how much – then of course… what takes precedence.
One way that has emerged to help us, is Artificial Intelligence or AI. Ai will help us manage the data, but it is a learning system, and without significant and ongoing input, the AI has no information from which to learn. However, in this case, the issue isn’t one of not enough data – it is classifying and interpreting the data.
Limited systems such as those that analyse sites and product yields, and even generate 3D forms, already exist and are in use. Archistar AI, for example, can analyse sites, and is already used by real estate and development companies to identify profitable investment opportunities – but these systems don’t tell us how the public feel about that. It doesn’t set out development guidelines, it just reacts to already existing compliance guidelines, be they good, bad, or simply just there.
There are certainly other AI systems in development. Personally, I’ve been working with an AI developer for years, and we certainly haven’t yet cracked the problem. We’re able to build small systems that tend to grow through their own learning, but prioritisation and human understanding is still the key.
One of the issues with mass amounts of data, is that the sheer amount of data can seem to give it complete authority, in terms of the brief and what the issues are.
We need only look at the home of mega data and data mining… social media: It can very easily just give us more of the same (like the echo chamber people refer to regarding Instagram and Facebook), or it can be manipulated by clever campaigners, even bots and malicious actors. A further consequence can be to exclude minorities, as they simply don’t produce the same amount of data due to minimal input points.
I give the example of my home country – for the last 200 years most planning decisions have been made using an almost entirely European model, and by those educated in precisely that system. There was, of course, an entirely different interaction with the land, and a deep understanding that had been employed on the continent for thousands, rather than hundreds, of years – I am talking of the First Nations people’s concept of ‘Country’.
In recent years, there have been great strides in recognising ‘Country’ and the traditional owners, and ever so slowly it is creeping into real solutions developed by urban theorists and practitioners.
One may argue, that as the First Nations people represent less than 3.5% of the Australian population, their voice would be lost in the input data, especially when the designer is overwhelmed with so many other inputs. It’s not that they don’t care – indeed, now that this has been brought to their attention, there is almost a rush to work in this exciting new (actually… extremely old) paradigm.
My point is, we (as designers) cannot just rely on data, or the sifting and collation systems and even AI assisted systems, as those still rely on the quality of the input, and by their nature, the quantity of data overrides the quality of that data. We still have a responsibility to cast a critical eye, to make connections and drive innovations, but to do so with the most open mind possible, and with our eyes on the prize… the quality of urban space for the end user, and their sense of inclusion both with the physical environment and with the process.
So where do we go from here?
Is Open-source Urbanism a pipe dream or just another buzz word to make us sound ‘cool’? Is it akin to greenwashing, in that (in some cases) a lot is said but little is done?
I don’t believe that.
To see where it is going, I go right back to the start. Open-source Urbanism, despite its complexities and challenges, is simply another method of approach… or thinking strategy. It is something we, as urban designers, developers and collaborators can learn from, be inspired by, and improve our work through… in the same way as those CEO’s I spoke about turned to design thinking, as a new way to drive decision making.
If you look back to the early days of open-source coding and programming, you’ll note that Microsoft saw the movement as a huge threat to its business, with both Bill Gates and Steve Ballmer speaking out on numerous occasions against it. However, since 2010, they have embraced open-source… starting slowly by opening some of their own source code, and reaching the point where, in 2018, Microsoft acquired GitHub, the largest open-source platform (or host) today. Now… they credit open-source as driving growth in their business, both in terms of product offerings and profit.
I think at this stage, a lot of designers do fear the open-source model – by democratizing the design process, they fear the role of the designer is diminished, that the soul of design… the poetry if you will… is lost, and that design becomes a paint by numbers, ultimately spat out by an AI assisted process that negates our years of experience and practice. Again, there is no need to fear this. We should learn from Microsoft and others, and start to truly tap into the potential of collective knowledge out there… of being open to behavioural patterns and new ways of learning about the communities we live in and design for.
Firstly, I think as a community of professionals invested in the continuous improvement of the urban environment (at all levels, from sustainability to community engagement, to health, to traffic management, to air quality and other metrics), we can acknowledge the importance of democratising the design brief, and the review process, of urban design… via Open-source Urbanism.
We can recognise the importance of every voice… of every story.
I believe we already implicitly know that design teams are fallible and finite, in terms of manpower and research capacity. By opening ourselves up to an open-source input, we can develop new and better solutions to the ones we have at our disposal right now. By feeding our outcomes back into the open-source urban design community, like a github for urban designers, we contribute to the source code… to the knowledge bank, and we can improve certain responses, or develop further understanding of community needs.
We have new tools at our disposal, that prior to digitisation weren’t available, or even conceivable. Thirty years ago, would you have started a conversation with a chair to ask how it is used? You’d be a madman! Now, you can literally do that… through smart sensors and other tracking and analysis tools.
We can use AI to overcome some of the information overload, and then overlay that with human empathy and design knowledge and thinking – to inform our process and, ultimately, produce better cities, better communities, and more connected social infrastructure… both in the tangible and intangible aspects of urban design.
At the end of the day, this is why we do this – to design and produce better spaces, that are more aligned with the communities they serve, and to respond and adapt to changing community desires.
These are the challenges – it starts with a mindset… a demand for change, and an understanding of what is available to us, then an earnest exploration of next steps. For me, this is the exciting world of Open-source Urbanism that we find ourselves in.
Speaker: Matthew Young