The last decade has seen the emergence and dominance of LED technology in lighting. From the beginning, I was always curious about three things: the strange, at times bizarre focus on product lifetimes; the need to rebuild the electrical infrastructure rather than simply patch the new technology into it without significant modifications; and the tension between modularity, where components are switched freely and often, and “integrated” fixtures, which ostensibly provide “future proof” and “seamless” solutions to lighting systems. This curiosity led me to develop a talk which I will give at LightShow West on September 25.
My experience in the earlyish days of LED commercialization, say circa 2012, was that at the time manufacturers were still concerned with proving the viability of the technology. One of my first hands-on experiences was testing a range of competing bulbs with different drivers and dimmers. The level of performance was mostly awful – one or two combinations actually blew up my testing board. But in the years before that, manufacturers desperately wanted to reach a lumen output that was adequate for most general lighting purposes, then to claim the dramatic efficiency that LEDs promised (what we now understand as real lighting quality wasn’t quite on the table yet.) They were keen to line up as many perceived value propositions as possible, and one of them was lifetime, since efficiency didn’t count for much with consumers then, and still may not. They competed fiercely on “specsmanship” – longer lifetime, whether real, practical or desirable (more on this below) must be better.
The unintended consequence of extended lifetime (inevitable in retrospect of course) was that it would ultimately destroy the fundamental business model of the lighting industry that had held steady for generations – steady sales based on a varying degree of planned obsolescence and repeated maintenance of installed systems. We’re now still dealing with the fallout from this.
I’m investigating the role of planned obsolescence in lighting and am up against what might be called the complexity of multivariate analysis problem- in other words there are many factors impacting the outcome of this, and the relationships between the factors are complex, so basically even in retrospect it’s difficult to tease out exactly what happened. But once I do, and I think I’m close to some useful conclusions, I hope to be able to apply this learning effectively to our current situation.
Early lighting systems, like Edison’s installation in Manhattan served by the Peral Street generating station, were sold as end-to-end solutions to the rich people who could afford them. They involved either on-site generation from coal gas (expensive, quirky, and complicated), or generation from plants located within a one-mile radius of end use. Lighting was the initial use for distributed electrical systems, so some of the bugs got worked out through lighting systems first, then other applications of electricity ( like home appliances ) followed once the electrical distribution systems were installed. Most relevant to this discussion is that if a lightbulb burned out, the power company sent a technician at their cost to replace it. You would think that the power company had much to gain from long-lasting bulbs, and you would be only partly right.
Several multivariate factors complicate the analysis. Bulb efficiency was impacted by several things, one of them being constant power, which was hard to deliver in early systems. Tungsten filaments which eventually became prevalent have a lifetime limited by physical properties that are hard to overcome – including the melting point of tungsten, thermal conditions, and the presence or absence of noble gas in the bulb. Bulbs also were more efficient, delivering more light, at higher wattages. Early power generation had the problem of load balancing on the grid, just as we still do today. Therefore, in the interests of the utilities responsible for power generation, more shorter lasting bulbs were preferable.
Eventually the business of bulb manufacturing generally became decoupled from that of power generation. The Phoebus Cartel was not the first of the cartels formed in Europe in the early 20thcentury to regulate the lighting industry, but it was the first to develop and employ the strategy of planned obsolescence, which seems evil to us now, especially with our awareness of the idea of the circular economy as described by William McDonough in his book Cradle to Cradle. But planned obsolescence eventually became a foundation of industrial capitalism, especially during the Depression, and a mechanism by which jobs could be created and the standard of living increased on a broad scale. Lightbulb companies realized that even if limiting bulb lifetimes to less than what was technically feasible may be not entirely desirable for end customers, they could make heaps of money with a guaranteed sales stream, and they proceeded to do exactly that. But it’s also important to consider that limiting lifetimes was not necessarily all about secret cabal pulling the strings behind the scenes, but had some real basis in sound engineering principles and the need to standardize quality – imperatives that remain relevant today and are playing out in different ways.
Understanding this complex history helps us to navigate forecasting and decision making today. A close reading of the history of lighting in the last two centuries shows that the same concerns surface repeatedly, including: lifetime, reliability, thermal management, glare control, maintenance, energy efficiency, sharing of distribution networks, reducing wiring, color temperature, nostalgia for eclipsed technologies, and more. Many technologies and designs were discarded, often more by accident than any other reason, and resurfaced later – a prominent example is DC power, which is reemerging for excellent reasons.
We see that the history of lighting, architecture, energy systems, and industrial economies are all intricately connected. The more we know about the interdependencies of each the better we can chart the near future. We need better ways to build viable scenarios for the future, and some of these draw on methodologies from the past, combined with the powerful and pervasive communication technology available to all of us today. I’m building some scenarios at the moment and will share them in my talk. Come hear me break this all down in detail on September 25, and if you miss that, I’m sure to have a paper out on it afterwards, so stay tuned.