On Tesla Powerwall, and the skirmish between Moore’s law and physical laws

My column for Dezeen from May 2015 explored the urban implications of Tesla’s Powerwall battery storage product. Here’s the original edit, first posted at cityofsound.com on May 15 2015.


The product launch of the early 21st century is a well-honed little drama. It’s become as streamlined and archetypal as the automobile launches of the mid-1950s, half-dressed girls running their fingers down the fins of a Cadillac; at least before Don Draper et al put that to bed in favour of the television age’s saccharine-soaked short films.

Now it’s the staged simulcast. It’s in California. A wide, deep stage, generally comprising black nothingness in order to foreground a giant video display and a single figure, a charismatic yet casually-dressed CEO. The CEO stands before a crowd of lanyarded acolytes, most of whom know exactly what they’re here for, yet feign surprise with a volley of whoops and cheers exactly on cue, as if press embargoes are hardwired into their tonsils. A few in-jokes, a few geek jokes, before portentous music heralds a well-crafted product video, loitering with intent over the sleek facets of a beautifully engineered object.

But last Thursday night’s launch was different. It wasn’t for an iPhone or a Hololens or a Chromebook. It was for a battery.

It was for a 130cm tall x 86 cm wide x 18cm deep box of lithium-ion battery storage. And yet it got the whoops and cheers accordingly. It’s just as well that the Duracell Bunny couldn’t see the attention this thing got.

Yet it could indeed change the world, this thing. It could indeed change the basic patterns of urban development, just as it could change the carbon footprint of entire societies.

Could it?

The CEO in question was Tesla’s Elon Musk, the CEO that makes Iron Man’s Tony Stark look like, well, Elon Musk. The CEO that, if he didn’t exist we’d have to invent him but oh yeah we did and that’s also Tony Stark. The billionaire genius CEO that delivered a well-considered, visionary and apparently humble keynote outlining the unbelievable promise of, well, a battery for your home.

Musk runs the now successful electric car business, Tesla Motors, as well as casually overseeing a list of more implausible lines of work, from a reusable spacecraft business, SpaceX, to Hyperloop, which can only be described as a theoretical subsonic air travel thing.

Musk positioned the Powerwall as the way to positively move our energy consumption away from fossil fuels, and their absurdly damaging generation and distribution models, and to shift our homes, transport and industry to renewable energy sources instead, principally solar energy.

The Tesla Powerwall is the first time that anybody has coherently and confidently aimed the storage battery at a domestic market. I mean really. There have been domestic storage products for years, but Tesla’s are the first to capture the imagination, to draw scalability and reliability from their electric car business, while halving the price overnight, with the entry level unit coming in at $3000. And available for order now. You can put this thing on your wall, connect to solar cells on your roof, and change the way that you generate, consume and pay for energy almost instantly.

The language is careful. The Powerwall offers “independence from the energy grid”. This does not imply leaving the grid altogether quite yet, and in reality, it remains nowhere near that for the average consumer, at least initially. But it would certainly enable you to use stored power instead when electricity prices are high, provide emergency backup during brownouts, and its modular extensibility builds over time.

This is not only better. The combination of rooftop solar and lithium-ion battery storage could soon be cheaper than the grid, too. The grid was one of the great inventions of the 20th century, but echoes the central organisation of that time, and increasingly seems a little out of time, rightly or wrongly.

For places with high energy prices controlled by handful of incumbent suppliers running relatively dumb grids, counterpointed with a decent and growing spread of rooftop solar, this could be genuinely transformative. There will numerous regulatory obstacles placed in the way by those incumbents, but it feels like history is on the side of the distributed solution.

Energy specialists, from journalists to industry figures, have been quick to query its value. Too expensive for mainstream consumers, they say, and that there are better technical solutions out there. Yet there is effectively nothing on the market to compare a Powerwall to, given the way that Tesla can market this product. They are positioning it outside of the existing energy sector altogether. It feels more like a Google or Apple product rather than — well, could you name another battery storage brand? That’s no accident. It also means all those energy specialists don’t really know what they’re dealing with anymore.

The dynamics of this new sector for storage have as much in common with Moore’s law as the laws of Ohm, Faraday, Maxwell or the original Tesla. The former, which posits (approximately speaking) that computing power doubles every 18 months or so, has held up for 50 years now, and is the creative engine behind the extraordinary influence that technology now has on contemporary culture.

This means that Tesla Powerwall needs to be considered as a version 1.0 release. The v1 is a beachhead landing; it gets the attention of consumers, investors, and sells. The v2 is what actually sells bucketloads, works well, cements a new product category, and consolidates entirely new markets. Pull out a v1 iPod, if you still have one, and compare with a recent iPhone. That’s the rapacious dynamic we’re dealing with here — what Erik Brynjolfsson and Andrew Mcafee note in “The Second Machine Age” is exponential progress — and finally applied to the energy sector.

While battery storage is not a pure software problem, as raw physics still define many of the basic conditions, that dynamic suggests that Musk’s intimations are not idly made, with Tesla’s algorithms determining when to switch to stored energy, how to optimise its performance. As Marc Andreeson has said, software is eating the world. Sector after sector. Here at last is a startup-driven product that eats an aspect of an unhealthy world — that of fossil fuel-powered energy generation — and potentially replaces it with a better one.

As such, it will shape cities too. Technology has always been the primary shaper of cities. Rem Koolhaas recently stated, when asked by the EC about smart cities, that “the city used to be the domain of architects.” Oh Rem. Perhaps his mouth was outpacing his considerable intellect there, as his Venice Biennale ‘Elements’ show explicitly recognised the fundamental importance of technologies as well as building components, just as Reyner Banham had patiently and brilliantly explained how building services technologies and engineering directly shaped architecture and urbanism, half a century previously. Architects have helped shape cities, of course, but can we really say that the city has been “their domain” any more than technologists, engineers and inventors?

Oxford University professor Steve Rayner has a good list of the technologies that have driven cities, such as the elevator safety mechanism and the flushing toilet adding up to skyscrapers, from air conditioning enabling a sprawl of cities and city sprawl, from the more obvious impact of the automobile to the less obvious impact of jam (as a low cost delivery vehicle for sugar to fuel the industrial revolution, and thus enable Manchester and all that followed.)

Powerwall, and what follows in its wake, will shape cities accordingly too, in equally fundamental ways. The most interesting questions about a new kind of urban design do not concern traditional architecture and urbanism, but instead ask how contemporary networked technologies change interactions, services and spaces in cities — Powerwall can be seen in this context. What kind of urbanism does it suggest?

  1. I was fortunate to be part of a temporary design studio in Helsinki a few years ago, which featured the Chilean architect Alejandro Aravena. He sketched out a model of cellular, distributed infrastructure for the new developments on the forested outskirts of Finnish cities; essentially off-grid, medium-density, using various renewables, and modular forms of what would now be called micro-transit, and so on. Nothing spectacular there, but still as rare as hen’s teeth in reality. Local storage like Powerwall, outside of vanadium flow batteries, has been the missing link in those sketches. Until now.
  2. In turn, this has a knock-on effect onto other centralised infrastructures — the tangled knot of cables, ducts, wires and sewers I alluded to in a previous column. With energy in mind in particular, and taken to its logical conclusion, it could imply erasing the step-down transformers, district heating plants, petrol stations that punctuate our streets, the cables draped across roads or in awkward ducts and pipes under the pavement, and further out, pylons, cooling towers, power stations, windmills and so on. Imagine the street free of all this. It’s unlikely to happen any time soon, given the insane energy loads contemporary society apparently demands, and the demands of mass transit and industry probably not catered for through local generation — but still. Imagine a city without that array of grid-based infrastructure, with the ‘fifth facade’ of roofs made productive, and energy stored near point of consumption.
  3. It implies a shift to lighter, more agile forms of infrastructure layers. You can design, and modify, these layers later, faster — they’re more malleable, adaptable. (Depending on another design layer: the ownership and legal structures.) They’re perhaps closer, again, to the dynamics of Moore’s law than the laws of physics, with all that entails. It implies an urban strategy currently being explored by those other v1.0s, Uber and Airbnb, of not building or un-building — optimising the existing urban infrastructure rather than expensively building a new one, running entirely new applications on the same hardware.
  4. I’d actually miss elements of that previous infrastructure. Will Self has noted the “irrefutable majesty” of pylons, these “giant humanoid figures” tracing connection across our country. I’ll miss the cooling towers at Ratcliffe-on-Soar, seen from the train to Sheffield numerous times in my life. But this is pure nostalgia, another potentially damaging fuel source we would benefit from weening ourselves off.
  5. This proximity to Moore’s law, and its subsequent lightness, also implies a certain flakiness, however. As startup culture begins to hit a city’s core infrastructures, we have to also ask questions about the expected levels of robustness, redundancy, security, ease of use, universal accessibility. We are surrounded by failures already, assailed by spinning beachballs or “Cannot join network” or “Unfortunately Maps has stopped”. Try to use the wifi on a British train. (Try to make a phone call on one, for that matter.)
  6. This is something the froth of VC funding can’t mask when it gets physical. While the celebration and acceptance of failure in startup land — there is even a failureCon — is important, it needs to be tempered as it directly merges with our physical world.
  7. I recall a great Jessica Helfand quote from the mid-1990s, when the term “information architecture” started to bed in; she said the main difference between information architecture and architecture is that the latter kills you when it falls on your head. That clean, meaningful division is fading, as digital and physical can no longer be prised apart so easily.
  8. Some of the outcomes could be awkward or inconvenient; some of them could be dangerous. Go and read this Golden Krishna article about using the BMW app to open a car door as intended. Then come back and ask whether we want these principles applied to the way the street works? To autonomous vehicles? To emergency services? To energy systems?
  9. There is of course a new kind of resilience through decentralised systems, an intentional redundancy through replication of small pieces, loosely joined — akin to what Adrian Lahoud memorably called a ‘post-traumatic urbanism’. Yet there is also, currently, a lack of robustness, of security, when it comes to using IP backbone for core infrastructure, or when not working through interaction and service design work cogently enough.
  10. And now the bar is being raised. Indeed, what is the domestic energy version of “Unexpected item in the bagging area”? I’m not sure I want to find out.
  11. Hardware is hard. Hence the name. Most smart home stuff has not worked partly because we mistrust its utility (although Justin McGuirk has brilliantly picked apart the broader political dimension to these new Internet-of-things-enabled designs.) There’s this lack of robustness, and a lack of intuitive affordances. Using a phone to turn on a light can rarely be right. As a form of architecture, Tesla Powerwalls will need to be designed with what Vitruvious called “firmness, commodity and delight” in mind. They’ve probably got two out of three; it’s the ‘firmness’ that will need to be perfect with energy. But at least Tesla’s track record of building a business in automobiles, a highly-regulated, highly public product that cannot fail (generally, anyway), suggests that they are better placed than most startups.
  12. An interesting follow-on question is how this might affect the psychology of urban communities. The off-the-grid story is traditionally associated with the lone woodsman, the Nordic summer house dweller, the beach shack — now it could be a semi-detached on Acacia Avenue, or a block of flats in Budapest. Will this withdrawal from the grid mean a similar withdrawal from civic society? As people lash together their own infrastructure, will they find it increasingly inconvenient to pay for the infrastructure of others, a basic tenet of living together in cities? A JG Ballard would unravel such a world beautifully; a kind of lithium-fuelled hybrid of his ‘High Rise’ meets HBO’s ‘Silicon Valley’ meets Felicity Kendall’s ‘The Good Life’. Equally, it might generate greater concerns for immediate environments. Again, it depends on the ownership model underpinning the infrastructure.
  13. There is indeed an implicitly Californian image here; both the promise of endless summer — Musk jokes about that “handy fusion reactor in the sky” — and the individualistic suburban dwelling model, with the Powerwall looking like its natural habitat is the multi-car garage that the people on TV have. It’s not like that in the rest of the developed world, and nor in much of the so-called developing world.
  14. Yet there’s no reason why a Powerwall has to work solely in a Californian context, or necessarily reinforce that NTSC suburbia.
  15. The German energiewende over the last decade or so has left the majority of the country’s renewable energy infrastructure owned directly by individuals and communities rather than traditional energy companies. This is one of the biggest stories in Europe; I never understand why it’s not bigger news.
  16. Yet the key aspect there is distributed ownership of infrastructure; local communities building, owning and using what they need — and that does play to the Tesla vision. Powerwall plus energiewende could be wunderbar. Connected arrays of Powerwalls — again, the v2.0 and beyond — could work for apartments just as well as the kind of monster homes we see on ‘Modern Family’. If we saw the terrace as a long apartment block on its side, there no reason why shared storage infrastructure wouldn’t work there, either. It might then imply new collective models of civic ownership; again, to borrow a German example, like the baugruppen.
  17. This is the key question about Californian tech; can we ignore their ideology and lifestyle and just steal their machines? I don’t see why not.
  18. Finally (for now), there is the greater question of the product lifecycle: about the material flows before and after. While any move away from fossil fuels is a move in the right direction, Tesla’s announcement also fuelled a debate in the ‘critical metals’ world about the amounts of lithium, cobalt and graphite in the batteries, and then in ground.
  19. Behind the development, a legion of extractive industries must gear up, their vast robotic trucks crawling across mines in Australia, Bolivia, Chile and Argentina, ‘feed stock’ suppliers for Chinese and South Korean processing expertise, and then on to Japan, and Tesla’s suppliers such as Panasonic. And soon, a new highway being built from the Chemetall-Foote lithium mine in Nevada, four hours south of Tesla’s Gigafactory site, part of an enormous package of state subsidies to attract the company there. (Note also, though, Shenzhen’s BYD, which has a similar ambition to the Gigafactory.)
  20. Musk claims in the keynote that such storage batteries could cater for all our energy needs. He admits it’s a fairly bold hypothetical. (He somehow balances apparently genuine humility and ambitious hubris as he does this.) Yet some doubt there’s enough lithium on our little planet to do this. (Of course, SpaceX, another of Musk’s ventures, sets up the possibility of extraction and retrieval from planets other than Earth. But let’s assume that one’s a way off, even for Musk.)
  21. The US Geological Survey says we have a fair amount in the ground and a fair amount as yet undiscovered. But still. The Gigafactory running at full tilt by 2020 would probably disrupt that assumption, particularly when competing with the needs of other electric car manufacturers.
  22. That is where the shared on-demand vehicle is far more disruptive — an innovation that actively reduces consumption, not just switches its fuel source, by using the dynamics of the so-called sharing economy. That would leave Powerwalls free to power the things attached to our walls, taking one particularly hungry beast out of the domestic energy profile.
  23. Yet interestingly, Tesla Energy is the exact opposite of a sharing economy approach, reinforcing that now deeply unfashionable idea of ownership. Amory Lovins once stated, “People don’t want raw kilowatt hours … they want hot showers, cold beer, comfort, mobility, illumination.” Musk is testing this thesis — will people care enough to see the connection between kilowatt hours and the cold beer?
  24. If they do, perhaps these Powerwalls become the energy equivalent of Vitsoe 606 shelving, people taking it with them as they move (and if Tesla they had one iota of Vitsoe’s customer service, that would be worth a few grand right there.) However, unlike Vitsoe, the tech dynamic means the Powerwalls could be redundant within a decade. It cuts both ways, technology, and perhaps indicates that the better model is ultimately rental, where the equipment is replaced over time, perhaps along with your solar cells.)

Taken together, perhaps this the most interesting emerging theme in urbanism? Networked systems as a connective tissue laced in-between existing infrastructures? It leaves core grid energy as supporting the heavy hitters of mass transit and large industry, with smaller users catered for by independent distributed energy, just as mass transit is now surrounded by legions of ‘micro-transit’ startups, working in the gaps, and new digitally-enabled participation models work below the radar of centralised bureaucracies. The physical forms, and interaction and service models, are modelled on distributed organisation. This is a new urban infrastructure: light, cheap, networked, optimising existing fabric rather than building anew. Yet also, individualised, fragmented, market-based, potentially throwaway, and with the internet underpinning it, and the extractive industries that power it, as increasingly centralised entities. Which is which?

Tesla cars themselves are insufferably dull objects. The forthcoming mass-market Tesla Model 3 has all the panache of a Volkswagen Jetta. Playing into a highly conservative market, a Tesla car is never going to described as the cathedral of our age, as Roland Barthes once said of the Citroën DS. The Powerwall itself is an entirely obvious object. One’s tempted to ask what would an Ettore Sottsass would’ve done with the brief; but that would be missing the point.

Design itself has moved on. It’s the design of the largely invisible and infinite crystalline network structures of interactions and services, or the planetary-scale manufacturing and supply chains that drive Tesla’s business, that could be seen as cathedrals, or perhaps even something closer to the mystical forces that cathedrals were built in thrall to.

The scale of ambition, rather than the scale of the object, is the impressive feature of Tesla’s thinking under Musk. While the fuss is over the design of an object — and if the Powerwall embodies a mainstream movement away from fossil fuels, then there will be no more alluring sight on earth — it’s the system design, the more fluid layer of services that is overlaid onto our existing infrastructures, that is the truly transformational possibility, predicated on this increasingly intriguing skirmish between Moore’s law and the laws of physics. That form of design, rather than traditional architectural thinking, is the force driving our cities now, just as previous generations of technology did.


A version of this first posted at Dezeen on 15 May 2015, with the original edit posted at cityofsound.com where you’ll find more like this.