It is three months now that offices in the UK have been closed due to the Covid-19 pandemic and we are slowly reaching the stage where lockdown restrictions are being eased and organisations are pondering and planning their return to the office.

Throughout the past weeks, we have blogged about various Covid-19 related issues and commented on the uncertainty we face, the need to factor in the social role of the office, the complexity of the task facing frontline facility managers and the pressures to balance contradictory needs.

There is one particular question that is close to my heart as a scientist and that has not received much attention here or elsewhere as yet, and that is what opportunities an existing layout holds for circulation planning. So, here are my views on configurational thinking and how to make the most of your current workplace layout.

It is always useful to review what others have proposed regarding layout adaptations. The typical responses from architects, designers and consultants come in two categories.

Firstly, we have seen plenty of computational simulations assessing existing floor plans with the help of generative design principles. Generative design, also called parametric design is a process by which an algorithm is fed with constraints and then outputs a large number of possible solutions. In the context of Covid-19, this means defining any parameters such as the 2-metre distance rule and then analysing where bottlenecks occur, or how many people could be ideally placed in a space. Examples include more basic offerings from Buro Happold or Gensler, but also more elaborate simulations by matterlab and Autodesk allowing the flexible adaptation of parameters.

The second category of responses aims to rethink the whole office environment, both from a layout point of view but also tackling behaviours. The most discussed response was probably the 6 Feet Office, proposed by Cushman Wakefield in the early days of the pandemic. Ever since its publication, the concept has been widely critiqued for creating rather soulless places, begging the question why anyone would want to return to such an office in the first place. A more recent example is the summary offered by Shape, which again thins out occupancy with 2 metres of distance in mind alongside the operation of a tight cleaning regime and behavioural adaptations such as filling meeting rooms with the most distant seat first. This concept is interesting since it contains quite a few ‘if possible’ recommendations, for instance introduce one-way movement if possible, or ask employees to leave through a different exit if possible.

It is exactly this space of ‘if possible’ that I want to explore further here. How do you know what is possible, when you look at the ordinary workplace layout of your organisation and find something that might not be as glossy or shiny, or ordered as the above designer-y solutions?

This is where configurational thinking comes in. Configuration is the way in which elements are arranged to form a connected whole. It tells us how corridors are connected, how staircases link floors, how separate rooms (such as meeting rooms or other enclosed spaces) are accessed and how open-plan spaces are linked in with the rest of the spatial system. Based on a research theory and method called space syntax, which was pioneered at University College London in the 1970s, configurational thinking tells you about the circulation routes people will use to move around the office.

The configurational approach of space syntax considers space as a network of connections. Based on the nature of connections and the resulting overall network structure, it is argued that different affordances and opportunities arise, for instance for how movement is distributed, or how likely it is to meet other people and where that is bound to happen.

Typically, space syntax research in offices would consider the beneficial impacts of face-to-face encounters, however, in the context of Covid-19 and with distancing in mind, the same principles of layout affordances can be applied.

To demonstrate this, I have created four archetypical configurations, each a 3 x 3 room structure (see below) inspired by a very similar example Bill Hillier shows in his seminal book ‘Space is the Machine’ (page 21). In this case the four structures have different implications for how distributed or concentrated movement is likely going to be; how controllable or predictable movement is going to be; and how likely encounter is.

I am also applying Hillier’s space type analysis (page 250ff.), distinguishing spaces by the types of connections they have: A spaces are dead ends; B spaces have one entry and one exit; C spaces lie on one ring; whereas D spaces lie on two or more rings.

Four layout alternatives of a 3×3 grid creating distinct affordances for movement and encounter

While appearing similar at first sight, the four 3×3 layouts have very different underlying structures.

Configuration 1 consists of a linear system with a majority of B-type spaces. This means they will have to be traversed in order for people to reach all spaces. Movement in case 1 will be rather distributed (7 out of 9 spaces will be used for through-movement) and people will encounter others rather heavily, as there is only one way in or out of the system.

Configuration 2 is much more tree-like with a majority of dead-end A-spaces, which can be relatively secure places in terms of distancing and minimising encounters. However, movement will be very concentrated in the B-spaces forming the ‘trunk’ of the tree, again leading to plenty of encounters as people move in and out.

Configuration 3 presents multiple loops with a majority of C-spaces. A system like this enables one-way traffic, as one route could be dedicated to in-movement while the other one is reserved for out-movement. Distancing would be relatively easy in this layout.

Configuration 4 is a grid-like structure with lots of choice at each point, allowing people to weave their own way through the layout. This means movement will be most distributed, but also more organic and unplannable. Encounters will be very likely as there is no clear guidance for movement flows.

In times of the Covid-19 pandemic, a layout with a combination of C-spaces alongside A-spaces would be ideal for natural distancing. Organising the circulation as C-spaces and spaces for occupancy as A-spaces brings together the advantages of both in terms of keeping people apart.

With this in mind, organisations can look at their own workplace layouts and begin to understand a little bit better what might work for them. In grid-like layouts, organisations may want to block off certain routes, thus eliminating choice from the system, which will make movement more plannable and circular. This would turn D-spaces into C-spaces. In a linear or tree-like configuration it might be possible to move furniture such as desks around to create more loops. Such an intervention would be relatively easy and could making distancing more natural.

A word of warning is due, however, and that has to do with the still emerging science of Covid-19 transmissions. There are many things we do not yet know reliably enough in terms of the relative risks of certain behaviours and how exactly they result in actual infections. I have been following news from the different scientific communities relatively closely, from epidemiologists to public health experts, social network modellers and ventilation engineers.

What we know is that the main transmission route of Covid-19 is airborne, yet we don’t know exactly what role touch and surface transmission plays. We know that the virus can theoretically spread through touch and contaminated surfaces, but we do not know to what degree that leads to infections. We also know that indoor environments are at much higher risk than outdoor environments and that air turbulences and air exchange are crucial, yet we don’t know exactly how to ensure proper ventilation to minimise transmission risks. We know that proximity and duration of encounters increase the risk of transmission (as for example shown in this analysis of a Korean call centre), yet we don’t know exactly how close is too close, or how long is too long.

The following example might help bring to life why we need to follow the emerging science and new insights very closely. The loop layout suggested above would help minimise the likelihood of chance encounters in corridors. But of course, one-way circulation systems create inconveniences, so the question is whether it is worth the hassle. One might argue that a brief fleeting encounter between two people passing each other in the corridor, even if they were closer than 2 metres (but let’s say, only for 10 seconds) might create the same infection risk as breathing in the same air that a colleague has walked through a minute beforehand (and possibly had to cough, or spoke with someone). These are subtle questions, but they might make a difference. The difficulty at hand is that scientists cannot study this in an experiment, because it would put people at risk, and hence would be considered unethical.

So why bother with the layout considerations I’m suggesting here? I would argue that it is certainly not worth refurbishing whole offices into 6 feet institutions, or to revert to cubicle hell. The transition period where we have to find ways to live and work with the virus (until a vaccine is developed and rolled out at scale) could be anything from six months to years. Being able to adapt an existing workplace layout in gentle ways to make it safer for employees will be a crucial exercise in the short term.

Configurational thinking will help with the appreciation of the potentials and opportunities of a current layout and will allow adequate responses as soon as the epidemiological evidence becomes clearer. And if more people understand the underlying principles of how layouts function by bringing people together or keeping them apart, this insight can be used not just for the transitional office, but also in the longer term to re-establish offices as social places for sharing.