By Clifton Stanley Lemon
This is the second post in a four-part series on Lighting and Innovation. The first post can be read here.
It’s supremely ironic that we are, ostensibly, in a state of stagnating innovation in lighting today, given that the universal symbol of innovation and inspiration is…the light bulb. Of course, the enduring narrative behind the symbol is a dramatic oversimplification – viral narratives typically are. Most of us know that Edison didn’t invent the lightbulb. We often fail to remember what he did invent – an entire system for local electrification, essential to the success of the new technology, electricity. All innovations that scale are the result of constellations of many interconnected factors, not necessarily technical and not necessarily even new.
We focus on single innovations like the iPhone (which we must talk about that more than any other single product) with almost complete blindness to everything that proceeded them, largely because that history is prohibitively computationally intensive and won’t fit into a neat viral-ready narrative. The simple, smooth magical little slab we carry around is comprised of an unimaginable amount of different complex technologies (well, 34 or thereabouts, specifically), as described in this graphic. One of my favorite stories about the interconnectedness of everything is this 2017 LinkedIn post about ox-drawn Roman carts impacting rocket launches.
Let’s start by looking at how innovations evolve in the context of what we often call “stacks” – combinations of technologies or business models.
A Story of Three Technologies
The highly simplified graphic story above compares commonalities and differences among three dominant technologies in the 20th century. The adoption and scaling of each were based on decades or centuries of prior technology development, invention, and innovation and only happened in combination with many different innovations, technical and non-technical (I have somewhat arbitrarily reduced the numbers of elements in each “stack” to two or three crucial ones, in fact in each case there were many).
Common Origins in Market Need – Each innovation started with a market need to replace or augment existing technology. With automobiles and electric light, the need was immediately recognized, but with the telephone, it was not necessarily so (below, I examine the differences between these two conditions). With the automobile, replacing horses became a critical need as streets in busy city centers filled with manure every day, an unsustainable urban condition. Electric lighting was seen as necessary to counteract the significant indoor pollution created by existing gas lighting (although exporting the pollution to the local coal gas electric generation plants required to power lighting merely displaced it temporarily). In the case of the telephone, however, rapid long-distance communication had already been achieved with the telegraph, and it took direct experience before people understood the transformative utility of real-time voice communication. In all cases, innovations began when someone somewhere dared to pose questions like “does it have to be only Morse code? Why can’t we talk over the wires?” and “what if we put that engine on a horse cart so it could drive itself?”
The Role of Stacks – With the automobile, the most salient technology, of course, is the internal combustion engine, but mass production and scaling were only possible with Henry Ford’s innovation of the assembly line, which was not called a “technology” then, although it probably would be today. The telephone (once called the “talking telegraph”) benefitted from a large existing network of wires installed for telegraphs – not a single technology per se – and eventually from a social innovation, the telephone exchange, which enabled people to reach the party they needed to call. While the inventions of Alexander Graham Bell and Antonio Meucci, and others, such as electromagnetic transmitters and receivers, were important, they were not singular inventions that resulted in the innovation of the telephone by themselves. Electric lighting has a particularly interesting history with technology stacks. Thomas Edison realized that in order to scale the technology of the filament light bulb, that he had not invented singlehandedly but instead perfected for commercial use, he needed to reinvent the entire idea of electricity distribution. Indoor electric lighting at first did, in fact, make practical use of existing gas lines for gas lighting, but beyond that, the entire system had to be built from the ground up: power generation, distribution, connecting hardware, voltage, and load management. The adoption path for lighting, in fact, had a unique twist – because it was the first commercial, residential use of a new technology, electricity, it laid the groundwork for the evolution of the electrical grid, as both industrial and residential buildings soon began to develop electrical equipment that could run on the electricity originally provided for lighting. If history has symmetries, we might imagine that innovations in lighting today around integrated building systems could have a similar effect. I’m fond of pushing this point of view!
The Role of Networks – Each innovation necessitated a network or a combination of different networks, and their nature depended on the nature and needs of the specific technology. The automobile grew and spread with the development of hard top roads and a national fossil fuel industry that controlled mining, refinement, production, distribution, and of course, ran on the road network. The telephone required an increasingly complex and far-reaching network of wires and exchanges that was built on the bones of the telegraph network, which ran along railroad rights-of-way. Lighting required an electrical grid, which began at the local level with DC power and gradually grew to the largest machine on the planet. All of these networks are interdependent and evolved organically – except for federally funded interstate highway projects, there was little in the way of centralized planning for the telephone or electrical grid.
Failures and False Starts – These two phenomena are different but related. In innovation, failure is the rule rather than the exception – like evolution, where most new organisms generated by genetic mistakes don’t survive, most new ideas fail. But also, like evolution, important changes would not arise without a steady stream of failures that produce a small number of variations that inevitably succeed. It’s important to put failure into its proper context, though – Silicon Valley disruption culture has lately come to fetishize failure and made it into a certain badge of courage: if you blow through billions of dollars of other people’s money with nothing to show for it, preferably several times, you must be a genius. False starts, on the other hand, are innovations that solve a problem for a limited time or range of conditions but either create more new and unforeseen problems or simply fail to scale according to their projected potential. In lighting, CFLs for residential use is a good example of this.
Shelved and Resurrected Tech – Innovation is a much more random process that we typically realize. The history of technology and innovation is filled with examples of inventions and discoveries that were put aside and revived later, or especially, inventions created in the quest to solve one problem that accidentally ended up solving a completely different problem. Some examples close at hand are the resurrection of electrical power in automobiles, which was an early contender during the development of the first cars. In lighting, Edison’s original platform, DC power, seems poised to make a significant return for many very practical reasons, including reducing conversion losses. Filament lamps also have made a massive comeback, quietly amassing a huge share of installed base in just a few years, although the LED filament lamps are a considerably different technology than the filament bulbs of the late nineteenth and early twentieth centuries.
Relative Pace of Change
It’s often observed that lighting technology has made dramatically fewer advancements over the last century or so than many other technologies. This may be true if you limit your perspective to consumer electronics and communications technologies. In the building and construction industries, there have been even fewer fundamental innovations in materials and HVAC equipment design in the same period. But a closer look shows that each innovation – the auto, telephone, and electric lighting – once established saw relatively few innovations with very large impact, although their trajectories are all unique. How much has changed fundamentally? Automobiles are still mostly internal combustion engines on steel chassis powered by gasoline. Telephones are still point-to-point communication devices running on very large complex networks, although they’ve acquired a huge number of additional capabilities. Electric lighting is still individual luminaires, made from metal and plastic, often with replaceable lamps and connected to electrical systems, completely separate from other building systems. Lighting controls are just beginning to evolve beyond basic dimming, but a vast majority of commercial buildings (on the order of 80%) still do not have multi-level or dimming lighting controls. Despite powerful prevailing tech narratives, and especially in the built environment, we should stop stigmatizing incremental innovation and prioritizing large “disruptive” technologies. To begin with, there aren’t many fundamental, game-changing innovations, and the building industry has all the disruption it can handle on economic levels alone – look at the sudden exposure to stranded assets that commercial building owners face today in downtown San Francisco, for instance
The graph above, from a 2013 Harvard Business Review article updated in 2019, is one of the most widely cited comparative pictures of the Faster Change narrative. But with a closer reading of this chart. the idea that change is happening faster all the time (see the Red Queen Syndrome in Part 1) doesn’t quite explain everything so neatly. The first thing I notice is that technologies with steeper adoption rates tend to be information and communication ones; for instance, radio has a curve in the 1920s that’s comparable to internet and cellphone curves almost a century later. Next, with the exception of the automobile, all of the technologies compared are based on electricity. Consider the effect of the highly unequal spread of electricity over the twentieth century – poor and rural areas had dramatically slower adoption rates. There is a similar inequality today in internet penetration between rich urban areas and poor rural ones – even in 2005, internet penetration was only 60% (today, the national average is 85%, but in rural areas, it’s 75% or under). Also, consider the relative prices of each technology. Automobiles are by far the most expensive consumer item on the list, and their slower adoption reflects the economic impacts of the Depression and both world wars. So we might conclude that white goods basically followed the electrification adoption curve, and once that leveled out in the 1950s, communication technologies had a faster adoption curve than white goods. Even if it seems like things are moving faster today, in context, they’re not necessarily.
Maintenance- the Eternal Stepchild
It’s hard to argue that the developers of any consumer product, building, or system routinely design for ease of maintenance. Financial incentives and risk management concerns for contractors dictate a quick exit from the building once the certificate of occupancy is delivered. David Edgerton sums up the situation bleakly but accurately:
“Just like land, buildings, and people, things have to be maintained, repaired, and looked after, often for very long periods. Although central to our relationship with things, maintenance, and repair are matters we would rather not think about. Mundane and infuriating, full of uncertainties, they are among the major annoyances surrounding things. The subject is left in the margins, often to marginal groups.”
In the past, most lighting professionals had little reason to contemplate maintenance, and today they have even less. The dramatically extended lifetime of LED lamps and luminaires can even exacerbate problems in design for maintenance, as when installing luminaires in twenty-foot ceilings without a plan to service or re-lamp them. Lighting control systems, however, face a much more complex maintenance problem, as the initial effort after installation needs to focus not so much on the hardware and equipment as on the people operating it. Most building control systems fail because occupants and facilities managers don’t know how to operate or maintain them.
A few lighting manufacturers have done some incremental yet important innovation in their products, like making agnostic drivers and luminaires with hardware that simplifies installation and servicing but there is much more work to be done. Controls manufacturers talk about “intuitive” designs, but controls remain as complex and infuriating as ever.
Even if you take a breathless techno-optimist view only, there are huge opportunities in maintenance. Sensor and analytic systems that are emerging on the substrate of lighting systems present sophisticated capabilities for predictive maintenance, extending the asset value of existing buildings and improving user health and experience. What’s missing is positive narratives about the value of maintenance in the first place, beyond “mundane, infuriating, and full of uncertainties.”
Build it And They Will Come vs Ask Customers What They Want
These two competing narratives are frequently invoked to explain innovation. Steve Jobs was the chief proponent of the Build It And They Will Come narrative, and he quoted Henry Ford in citing the futility of asking customers what they want. The background of these narratives shows that they’re not necessarily in conflict at all and that they don’t represent a useful dichotomy. Jobs built a very strong culture at Apple that was completely demo-based. Rather than rely on a lot of research, it was much more productive to establish a clear design brief then go to the lab and mock stuff up, over and over, until you got it right. This was also basically Thomas Edison’s approach, and there’s a lot to be said for it. But understanding the market need, however you do it, is key to success.
You need to ask the right questions and solve the right problems in the first place – this mindset is an essential component of innovation and of good design. As an inventor or innovator, it’s often very difficult to know exactly where you land along the continuum between these two narratives, and it’s much easier to understand success in hindsight when it seems obvious. George Eastman did not know for certain that people wanted or needed portable cameras when he invented them, so he went about starting photography clubs and socializing the idea until the portable camera suddenly exploded. That’s a historical example that supports the Build It And They Will Come narrative. Perhaps the developers of self-driving cars can be forgiven for allowing this narrative to propel their relentless pursuit of what they assume is a certain future, but at the moment, people don’t really need or want self-driving cars, and the innovation looks like it’s headed for the False Start exit. We do need to ask what specific problems they’re trying to solve. At least electric cars, in large part, are being promoted as a way to facilitate decarbonization, but cars, electric, self-driving, or both, are still…cars with huge, pervasive, and increasing environmental costs. Does the world need more automobiles? Two-thirds of world population experiences severe water shortages at least one month a year, yet there are 1.5 billion cars and trucks on the planet.
Vaclav Smil, in Invention and Innovation, A Brief History of Hype and Failure, says,
“As in the past, we will succeed in some quests but fail in others, and we will not be able to ignore the fact that many gains will take place within limits rather than being the products of unlimited progress. We should restrain our ever-present compulsion to forecast how new inventions will shape our future: retrospectives of such efforts show only very limited success and a preponderance of failures. A better, safer, more equitable world will require many truly transformative inventions, but we will know the extent or absence of these expectations only when looking back – and we must hope that some of the items [we dream about today] will become realities before the middle of the twenty-first century.”
While we don’t need to understand everything about any particular technology in order to use it and even to innovate with it, it does help to see it in context – historical, technical, economic, and behavioral – and to think beyond the single component to the larger system, which is exactly what Edison did. The larger system we need to consider in lighting today is different than at the turn of the twentieth century, but there are still some useful similarities. We’re back to considering DC power, for instance, and we could end up reinventing electrical systems in the service of making lighting better, although the inertia to overcome even the AC standard is considerable. Things change slowly in the built environment,
Another fundamental technology disruption, like solid state lighting may or may not be on the horizon any time soon, but we don’t need to wait for it in order to continue useful innovation in lighting. As Dr. Robert Karlicek told me recently in describing his research and innovations, “we just want to make lighting easier to buy, install, and use.” This simple idea and others like it are excellent starting points for innovation, as they recognize user and market needs and aim to address them.
We should probably stop worrying about “future-proofing” since we do such a lousy job of predicting and think about “present proofing:” making lighting easier to use with the technology we have on hand right now. We can start by looking at the larger problems we face, and energy management through smart buildings and grid connection is one area that lighting touches and can impact significantly. In order to effectively innovate here, we need to look at buildings and the grid as whole, connected systems. One of the ways to do that is to look at lighting as a whole system rather than a collection of components.
In Part 3, I will dive into a promising new platform for lighting systems developed by my client Quarkstar, one that considers light extraction, streamlining manufacturing, and truly exploring the inherent capabilities of solid-state lighting.