Many trucks driving in motion blur on the freeway at beautiful sunset. Rush hour on the motorway near Belgrade – Serbia.

What can architecture learn from logistics

Architect Jesse LeCavalier discusses what city planners could learn from Walmart, Amazon and other influential retailers

Read intro

An interview with Jesse LeCavalier

An interview by Mihailo Popović

A discussion about relevance of logistics for architecture and city planning with Jesse LeCavalier, an architect with special interest in logistics and retail infrastructure. Discussing the trends in city planning, their evolution and possible future paths.

Jesse LeCavalier is an assistant professor of architecture at New Jersey Institute of Technology and a member of Co+LeCavalier, a senior researcher at the Singapore–ETH Center’s Future Cities Laboratory. His work has appeared in Cabinet, Public Culture, and AD, among others. LeCavalier’s essay, 'The Restlessness of Objects' was the recipient of a 2013 Core77 Design Award and his article 'All Those Numbers' was named by the Atlantic as one of 'Nearly 100 Fantastic Pieces of Journalism' in 2011. LeCavalier has a Doctor of Science from the ETH Zurich, on retail logistics, a Master of Architecture from the University of California, Berkeley, and a Bachelor of Arts from Brown University.
Hide intro

What brought you to questions of logistics? Why Walmart?

It started with an interest in changing the nature of public space in the United States. I was interested in different kinds of sites of political engagement. For example, the shopping mall as a site of disputed public territory, where you need a permit to distribute flyers or to get signatures for petitions. That has been a debated site of public engagement. Different parties claimed different. The owner would say it’s private, but people wanted to use it as public. And there has been a number of supreme court cases about that. From there I was curious about how something like a Walmart store, which is one of the few places that is open for 24 hours in smaller towns became an urban hub. Then I was simultaneously interested in subjects of infrastructure, and how larger scale design operations have repelling effects, in terms of setting certain processes in motion. I realized that they are the part of a much larger network, but what really started to distinguish this company in terms of its capacity, was its relationship to logistical network and logistical expertise. Its awareness and deliberate investments in logistics as a site of potential competitive advantage.


UPS launches ‘We Love Logistics’ campaign, September 2010

Why is logistics interesting and relevant to contemporary discussions of urbanism, architecture and design?

We can think about early industrialization and what happened to cities with the growth of factories and that kind of material-​intensive investments in locations where goods would be brought in and consolidated. And I am mostly talking from an American point of view. In the US, for example, some places like the north-​east is where all sorts of trade routes brought raw materials into cities, and then those cities consolidated around giant industrial processes. You could think about Ford or General Motors as a paradigmatic company of early industrialization. Then there was the shift toward the information age of the IBMs and the CISCOs and things like that. I have been thinking whether we can call the age that we are in now the age of circulation, where industrialization and communication are brought together uncomfortably. I think that’s what’s interesting about this moment. If you look into an advertising image of logistics by companies like UPS or FedEx, they present it as an immaterial thing. It looks like a seamless process of movement when in fact it’s intense and quite heavy.

So it is basically combining the two: trying to deal with communication while simultaneously managing the physicality. What you get as a result is a weird duality between a tendency toward abstraction and concreteness simultaneously. Amazon and Walmart for example, the more they try to deny space by claiming that things can be delivered instantaneously, the more they have to encumber huge territories in order to make that happen.

The more Amazon and Walmart try to deny space by claiming that things can be delivered instantaneously, the more they have to encumber huge territories to make that happen

I have been trying to think about logistics at different scales and I’ve been trying to look at the Walmart story in basically three scales: scale of the body, scale of the building and scale of the city. Mostly I have been talking about the scale of the city and how these different dynamics are acting in the city. The other is the building scale. Walmart says it’s a retailer, but I think they are really a logistics company in a sense that mostly what they are concerned about is moving material around in a fast and reliable way. The way they think about their buildings differ: not as a capital investments but as operating costs, as equipment. They don’t see them in the same way previous corporations have seen their buildings. And this is true for any kind of large retailer like IKEA. What does it mean to be designing these kinds of buildings? What can we learn from the way these buildings are designed? It might suggest the opportunity to make architecture itself more nimble. I think we might be seeing another gap that needs to be filled in terms of these questions of infrastructure logistics.

Can you explain a bit more about Walmart’s attitude towards their buildings?

What’s interesting to me is how Walmart’s attitude towards buildings is not about preciousness of an object, nor architecture as enclosure, it’s really about architecture as interface and mediation. The building itself is a way to connect the shipping network to shoppers and their local sites. They are built to be flexible to adapt to a range of surprising conditions, to absorb the risks of Walmart’s operations.

And how is this new age of circulation taking shape?

The ability to automate information management has been part of this technological shift. The barcode has enabled instantaneous recording and monitoring of information in a way that was not possible before. The barcode is really dramatic because it allows different kinds of organizations to keep track of the movement of goods with all sorts of different data sets. In Walmart’s case, consumer purchasing behavior, linked to location and linked to season and linked to demographic trends, allows them to predict consumer behavior.

Are they really predicting consumption demands and desires so well? Because, as you showed, they need less and less storage rooms in their shops as result of sophisticated system that is feeding the demand?

There is a distinction between pushing and pulling the market. The old model of motor companies like a pre-​lean production car company would push their inventory to the market. They would make a certain number of cars and force their dealers to sell them. With the abandonment of lean production it became a pull model based on consumer demand. The consumer has a lot more power, because they are able to engage production based on their desire, but that’s also linked to the company itself nurturing and cultivating that desire.

It’s not so much that there is this autonomous consumer desire that Walmart is just responding to. There is a coproduction of that. They have a relationship with private weather companies that allow them to track weather patterns and respond more quickly. In US, the national weather service has a duty to the public to not be risky or alarmist, so they tend to be very conservative in their predictions and wait as long as they can before making an announcement. When hurricane Katrina made fall to New Orleans Walmart’s weather service, having the obligation only towards their client, was able to risk earlier prediction that it would hit land sooner. And they did that. As a result Walmart itself could mobilize more quickly than public services. There are these different kinds of scales based widening the gap between public services and private services. Because Walmart is able to develop their own set of various kinds of intellectual expertise, they are able to also operate in different kind of way.

Walmart’s attitude towards buildings is not about preciousness of an object, nor architecture as enclosure, it’s really about architecture as interface and mediation

The other example how they use their data is about how they choose locations, and this is where it intersects with urbanism. They have these marketing protocols linked to growth, saying where and how they need to expand, so they use processing software. In one they look at the locations that are currently available and how these compete with other locations. In the other one they do it in an opposite way. They say: ‘This is the kind of demographic we want to reach, so where should we build a building?’ If you extrapolate, there is implication that eventually building of cities becomes automated. There is an increasingly absent set of design choices.

picture 2

Walmart distribution center scheme. Designed to be flexible to adopt and keep Walmart’s operations going. Courtesy of Jesse LeCavalier

Through logistics systems we can see encapsulated future versions of cities. If location strategies can be optimized, and if Walmart does it well, there is a tendency to say we can do something similar. The idea of a city as something to be optimized should be taken seriously, but we should also be careful about it. We have to find ways to remember that a business is different from a city.

As we are talking about city and a building, I want to add one more thing, which is about the scale of a body. I think one of the reasons to look at logistics, is logistics buildings themselves. The large distribution centers and things like that, could stand in for a future and not implausible future environment, that we all may be inhabiting. One that is automated and incredibly data rich but also basically eligible to us. The more those systems become automated, the more we, who have to engage with those systems, are not equipped to read them. And so we have to get all these mediating devices to access this information. Those large data and distribution centers serve as a proxy for the environment that we might all be living in the future. So what future environments might be, as things become increasingly automated?

It is interesting what you said about Walmart buildings, that they can’t be seen as single objects, as they are part of one big thing. It in a way answers the rising questions of how more and more buildings that are built today have no local specificities and could be planted anywhere in the world. Should Walmart stores have local specificities?

One of the things that is interesting about Walmart is the means by which they build their buildings. They develop a series of prototypes and those prototypes are never built. The prototype describes certain configurations and set of requirements. When they have a new site and need to build it, they determine which kind of prototype works for that site and do this process called ‘site-​adopt’. The prototype gets modified in the field and becomes a different version every single time, even though it’s the same building. In a way it is locally adopted. It doesn’t have symbolic local relationships. They are basically copies without originals. They use that effectively because they can build them very quickly. WalMart is building several hundreds of stores a year, which means one opens every few days.

There other non-​geopolitically aggressive ways that building or architecture could be used. Any kind of corporate expansion processes is about securing territory and finding strategic location to ensure future growth. Is that capacity of buildings useful for other reasons? Is there something else that is good for that, other then getting locational advantage? I think it’s something to think about.

Could you say or speculate how logistical systems will look in near or more distant future?

The question is what’s the nature of logistical space. I wonder if you can connect it to these ideas of prediction and anticipation. So maybe we can talk about logistical space as being anticipatory. Maybe then communication space was reactive? Maybe we can think about future ideas of distribution that would have things that will people likely want, suspended in some kind of state of readiness. For example, you could imagine at any given moment there was a truck within one mile of you that has a book that you want, you order it and receive it immediately. That’s a different kind of attitude towards distance. It’s less about point to point and maybe more about kind of suspended state of readiness and anticipation.

You have probably heard about the KIVA company. Amazon recently purchased them, and it was the largest takeover since they bought Zappos – they spent 775 million on it. Kiva produces small robots that can bring the shelf to a person instead of having a distribution person do it. What that means for Amazon is huge, because there is a different kind of relationship to labor. It’s increasingly automated and the layout of the stores themselves start to conform to desires, so the more popular items stay closer to the people picking the orders, and less desired things get pushed and pushed further away. The building itself is changing. The previous model of distribution center is incredible materially intensive thing. Imagine 1 million square feet of conveyor belts and mechanized loaders and all of these things… And automation does away with a huge amount of that because all you need is simply the most basic enclosure. So the architecture becomes even more dematerialized. It requires the thinnest possible membrane to keep the weather out. I think that starts to suggest larger sets of tendencies towards suspension, towards anticipation.

When hurricane Katrina made fall to New Orleans, Walmart’s weather service was able to risk earlier prediction that it would hit land sooner

What happened to production itself?

Externalization, pushing costs further and further away. And a part of it is because they have command of information. They have proprietary software system that allows their suppliers to see what’s selling. So, if you are toothpaste manufacturer, you can see how many units are being sold and where but, Walmart can also see around inventory. As a result they can also push back and force other kinds of externalization. Sites of worst kind of labor are pushed to the places where it’s cheapest. And that brings up issues about humanitarian concerns. There is a recent book from Deborah Cowen called The Deadly Life of Logistics that talks about this.


Automation of site selection. One of two software models that Walmart uses. This is where logistics intersects with questions of urbanism. Courtesy of Jesse LeCavalier.

This is why I wanted to bring up this question of behavior. One of the versions of this story is, that those externalities will just keep going to the places where they are going to, and as they will then bring up quality of life slightly at each place. And gradually middle classes emerge, and gradually labor costs rise, and gradually that labor gets externalized elsewhere… And how far it would take to get to some point when that wouldn’t be possible anymore? What’s the limit of that dynamic? And will there always be other kind of cheaper source of labor, or there will be a point when it will loop back on itself? Or will it reach a point when automate labor will be cheap enough? And then how people will be living?

Amazon vs. Walmart? Which model will prevail?

It would be more interesting to ask what would happen if one or the other one succeeded? What would be the consequences of one prevailing? If some kind of hybridization occurs? For me, that would be the way I would try to answer it.