Over the past month, I had an interesting conversation with some big thinkers in the lighting industry about lighting and the Internet of Things. The conversation started with an email asking for experts and ideas for an article on the topic I’m writing for ELECTRICAL CONTRACTOR’s December issue. Here’s what they had to say, raw and unedited, though names have been changed (except for mine). Enjoy!
From JOHN:
Hi CRAIG,
I’d be happy to discuss as a specifier, person interested in the topic, retail grade system user and with a background in mechatronics. I’m fairly cynical, but I also have a sense of where opportunity remains, which might be of interest to electrical contractors.
From BILL:
Hi CRAIG,
I’d be happy to discuss from a high level technology point of view. For applications in the near term, I, like JOHN, am somewhat cynical. Longer term, though, IoT will increasingly drive the lighting/architecture design bus – though this is a complex topic with many technological and social dimensions. Of course, the skill set for your future electrical contractors will also be more advanced – more computer tech than electrician. Aren’t lighting companies asking for that already?
From JOHN:
Hi BILL,
In the interests of unnecessary debate, I thought I’d throw something in.
I agree that the control protocols will be increasingly IoT, but I think it’ll become less of a topic, not more. I buy any keyboard, and plug it into any computer USB, it works. On Apple, PC, Linux, etc. I think we’ll see that the communication means becomes ever less important in lighting (with attendant security concerns!). Before too long, lighting will be like office power – connect it to power, and one of several open protocol data networks (Bluetooth, wifi already dominant in consumer grade), whether wired or not.
Even if there are many protocols, we’ll see power supplies that are conversant in many or all (witness drivers that now run DMX/DALI/0-10V as native), so the hardware will handle it. (Oh to have royalties on a key code chunk!). But integrating them physically will get ever easier. And as the price drops, the cash available for services integrating will fall. I am ok paying 40% on top of a Lutron system for commissioning, but not on top of a $50 philips hue lamp!
So the skills level of electricians wouldn’t need to go up by much.
Configuring retail IoT lighting is pretty easy. But finding a way to charge for programming scenes, daylight harvesting, validating sensor positions, calibrations, etc, will become an ever larger percentage of the costs of a system. In turn meaning that it’ll come under focus and be the thing clients want to cut. I’d love to have a business paying college students to commission Philips Hue systems. They’re easy to do one lamp – but an entire house???
Anyway, IoT (or as I call it, Illumination Devices Internet of Things, IDIoT) is coming, and it’ll be cheap. It’ll be cheaper simply to select a generic control type power supply from a reputable manufacturer, and job done. When we get to that point is hard to tell – we’re almost there now with Alexa/Google Home at the retail level.
Gonna be an interesting few years until the open protocol is in place!
From BILL:
Thanks JOHN – a voice of reason, and generally I agree: Ultimately, advanced lighting systems will be so smart that any IDIoT will be able to install them – and they might even be self controlling and self-commissioning, and using all of that data to sell us more stuff! And just as I type this, BRUCE’s email comes in – more good points. I actually think we all agree on what the end will look like, but the question is – who will monetize any of this stuff? (BTW – Enlighted got bought by Siemens a couple a weeks ago – not too little anymore – and perhaps a harbinger of other acquisitions to come.) And now PHIL’S email …
I have to confess that I am looking at this from a lighting device/lighting fixture manufacture and wondering how they survive if they cannot derive revenue from the deployment of smart systems. In this context lighting will survive. If lighting is just a dumb plug-in to a smart socket (with a control interface dictated by the network or data aggregator), there isn’t a whole lot of incentive for continued innovation of SSL (or any lighting) technology.
The IoT/data aggregators covet what lighting has: ubiquity, vantage point and power (all for their sensors). If lighting companies give up that turf to Google and its ilk (or win what I call the battle for the smart socket), then I am not sure lighting companies have a very exciting future except for color tuning and dynamic beam steering.
Thanks all for the inputs (but email is an extraordinarily inefficient way to brainstorm) 🙁
From JOHN:
Adding to your inefficiency BILL, another email.
Largely agreed on everything. But I think the lighting industry needs to take a deep breath and remember that luminaires are not and historically, have not been anything but light sources with optics. Ballasts, etc, are a relatively recent thing to consider even “controllable”. Going back to the luminaire object being the sale would be a reversion to the norm.
Now, I fully concede that the consumables part changes – modules instead of lamps – but with the added sale of a power supply. All of which needs occasional maintenance. Less frequently than a lamp, but the revenue stream would be more than an annual 99c lamp if you’re changing a $30 component every decade…
So again, let’s take a deep breath.
There will be a lighting market into the future, and it may well be about producing good luminaires, with good optics, in keeping with trends on office working environments, aesthetics, health considerations, and just generally lighting shit.
From BRUCE:
In short.. EXACTLY.
The vast majority of lighting is pure basics; troffers, strips, cans, track, sconces. High-end lighting design, or even lighting designed to enable or incorporate IoT.. is the cream.
We tend to focus on the cream and ignore the bad tasting coffee below…. there’s a lot of money in that bad coffee.
From FRED:
CRAIG….look what you’ve started!!
From CRAIG:
Thanks for all your comments and insights!
One area I’m interested in is stratification. When we talk about IoT, often we’re talking about the most robust realization. The lighting market doesn’t work that way, nor the controls market. I’m curious about whether there will be IoT capability tiers based on building size or vertical markets, with lower tiers being introductory levels for more robust realization later. For example, OSRAM’s Encelium software today includes thermal mapping and shows occupancy patterns. Is this not IoT, call it baby IoT or IoT lite?
From BRUCE:
IoT in my concept is stratified by application. The challenges with IoT ‘FAT’ is that the number of sensors required to do all potential applications would likely collapse the financial viability.
I believe what we’ll see is a focused IoT to address 4-6 sensor applications (or less). While more could be added later.. I think the adoption curve will see a small, finite number of sensors.
From a sales perspective, I can envision a company offering a typical retail suite of IoT: asset management, space utilization, security, and ‘beacon’ access for incenting buyers.
Industrial would be a different offering, and could include temperature monitoring of equipment, occupancy sensors, air quality, etc.
Healthcare.. . different,
Each ‘typical package’ will be supported by financial justifications to the buyer (end user, or space ‘owner’) to support the expense and establish the analytics they’ll require.
Etc.
From BILL:
I agree with Ted’s segmented adoption version. ROI will be key to driving adoption, and there will be different use cases. Wild card will be a healthcare like adoption driven in commercial spaces with the evolving Wellness program – depends on how far Delos and the Well Living Lab (JV between Mayo Clinic and Delos) drive this. (The International Well Building Institute is sort of a front for Delos in my opinion, all of these wellness business parts are part of a well-oiled real estate machine started by Delos.)
From BRUCE:
And IoT is still highly reliant on the overall energy story to fund the implementation, so IOU’s are trying to come to grips with how they get involved in ‘connected lighting’. I had an hour long interview with a consultant for ComEd and we talked through the issues as seen through the eyes of the IOU; with demand side investments and whether they include IoT as a second case.. and will these systems make sense to communicate with smart meters… it was an interesting conversation for me.
From JOHN:
I keep coming back to:
Even pre IoT, the power supply/ballast/driver was a third party element. And so it shall be to the future. That will be IoT, the luminaire will not.
Just my 2c.
From CRAIG:
Just to throw some fuel on the fire, I wanted to share this item a lighting company executive shared with me for an article I’m writing about connected TLED lamps:
“The lighting industry is at an inflection point. Consolidation and attrition are the future for standard, unconnected lighting. Connected lighting as a standalone solution is self-limiting in value, expensive to set up and maintain, and closed, making it unacceptable as a long-term path. The best path for our industry is the best path for customers—an open-architecture industry standard that is simple and affordable to deploy; delivers the lighting, control, and energy management that a lighting system should; and transforms the lighting system into the breakout IoT infrastructure that enables true commercial value and sustained future growth.”
The gist seems to be: Offer lighting packaged with integrated, open-protocol controls in a connected lighting solution that can play in future IoT adoption, or you’re screwed. Thoughts?
From PHIL:
I’ve mouthed off elaborately about all of this before at length, but here goes…
I keep thinking more and more that the hype about IoT is fake news – forgive the use of such a poisonous term, but it’s true. One of the underlying problems with the way we see technology is that we assume that every invention is inevitably going to succeed because of things like Moore’s Law- tech companies and investors assume crazy, rapid cancerous growth and adoption because that’s what happened with computers, smartphones, and a few other limited areas of tech. But Moore’s Law is really anomalous- rapid improvements in technology rarely reach this kind of speed. And most innovations and inventions fail, so we worship “innovation” and ignore the more important things like craft, behavior, maintenance, and usability. Many very important technologies change very slowly , if at all. And we can cycle back to inventions that failed years or decades ago and implement them effectively. DC power is a wonderfully broad and impactful example of this.
Now it’s becoming viable because of totally separate improvements in and reduced costs of distributed generation, renewables, advanced storage, and smart networks. This is where IoT can have a much bigger impact.
I’m always amazed at the paucity of imagination when people create stuff like highly granular indoor positioning- when pitching what we could do with it, it’s always like ”we can send coupons for specials on Mayo in Aisle 8.” Like we all want more spam. Please. And huge distributed sensor networks can monitor the movement of people and things in buildings, but almost none of the data collected is being used for anything meaningful at all. Most inventions end up being used for a completely different purpose than the problems they set out to solve: we forget how accidental innovation and invention really is. A lot of the amazing new stuff we’re inventing is truly cool and might be really transformative, but we mostly don’t know what to do with it. We don’t have a practical application mindset at all, we’re perpetually stuck in the Genius Professor mindset.
Then there the problem of total lack of experiential reference. We’re spending a lot of time talking about adaptive lighting driven by AI, and now of us has ever really experienced anything quite like that because it doesn’t really exist yet. We’re still trying to deal cognitively with dimming. Intelligent lighting systems that know how you’re feeling? Sure we can build those, but we haven’t, and why would we? Designer people make most of their important decisions based on how things look and feel, because they know that’s what normal people do. But we spend very little time thinking about how normal people feel, what they actually want, and how they actually behave. We’re mostly in the mode of saying” Here’s this new tech- it will change everything, it’s inevitable, hurry up and invest in it or you’ll miss the boat.” The narrative of technology is driven by companies looking for really quick really fast financial gain, like all the Silicon Valley new media and consumer electronics companies.
I do think it’s a threat that IoT driven companies can encroach on lighting as a building service, definitely, but that doesn’t mean we can all forget about the business of lighting. Very few people are talking about using IoT to monitor actual lighting performance- delivery of basic light. It’s always about energy (we have to let go of efficiency- we’re already mostly there in lighting) and surveillance (we are at increasing risk on this score). No one is using IoT to deepen our understanding of behavior and light. We can watch people move around in a space but our data gathering and analytics give us very little explanation about motivations and causation, despite a decade or so of frothy blather about it.
I’m also really sick of hearing “were at an inflection point, adapt or die.” This brings out the inner Luddite in me pretty damn quick. We’ve been at an inflection point for the last 10 or 15 or 20 years depending on how you choose to slice it. Every year is going to be the game changer. We blithely allow the vacuous bloated self-heroic hyperbole of Silicon Valley to control the narrative in the face of much evidence to the contrary. We don’t need any more speed and relentless change – we need Slow Tech.
And to propose making building systems even more complex and difficult to maintain, by adding IoT into the mix, seems inconceivable to me. No one has ever been able to answer the question “who will fix it when it breaks?” adequately when I ask it. We’re losing our electrical industry building trade labor force, who will train the workers of the future in IT driven building systems? Because if all our AI and black-box seamless integration in the shining new Jetsons future is really fabulous and wonderful, it will be all that much more catastrophic when it really breaks and no one can maintain it. I understand the military is now training people for operations in a non-internet world, in case it gets taken out, and it certainly could.
From JOHN:
Wot ‘e sed.
From BILL:
Lots of great discussion, and this from the IoT and sensing guy. Just some high level thoughts from the mad scientist in the group:
• When LEDs for lighting first came out, most lighting designers ignored them – some thought they were just a passing fad. Now, LED can do just about everything and do it more efficiently, though adoption still lags due to cost and slow replacement cycles. The only real industry that got killed here were the bulb/tube companies, and now Asia rules that roost. Metal benders (my name for the luminaire companies) have just had to learn some new tricks for using LEDs instead of bulbs/tubes (and had to learn Mandarin to source LEDs and modules at the lowest price).
• LED disruption enabled a lot of high tech startups to look at driving new functionality into lighting (color tuning, sensing, LiFi) – lots of hype, many already closed or sold off in a fire sale, some still pushing. Increasingly, the fixture companies are moving into these spaces, but not with good early results (see Cree, Acuity stock prices, though they are making small inroads in building management efforts).
• Fairly granular occupancy sensing is becoming increasingly important for high efficiency buildings, and sensors in the ceiling (for ubiquity, vantage point and access to lighting power) is one attractor for some big DOE funding trying to hit 30% HVAC energy cost reductions (I know – I just got $2.3M on lighting based occupancy sensing for commercial spaces integrated into HVAC control).
On a separate note, big realty companies managing are the ones who are mostly interested in this – I saw a presentation for a “smart building” company monitoring occupancy in a high rise building show that their system was so good that, during the partial solar eclipse in August 2017, their system sensed a two hour dip in occupancy when the folks went out to see the eclipse and their software automatically rebalanced HVAC and lighting loads, saving a small but not trivial amount of energy costs almost by accident. They are big on IoT in the ceilings…
• Depending on the trajectories of lighting and human wellbeing (or HCL – the term we are trying to avoid) and LiFi, IoT and data will be connected to lighting, as controls will be too difficult for mere mortals to operate. IF this comes to pass, the metal benders will have difficulty separating themselves from IoT and putting the brains in their products because there will be value and revenue there. Alternatively, if they want to stick to their knitting – the IoT guys will own the smart socket and the smart pole and tell the metal benders what their plugs need to have if they want to make a sale.
I’ll end with Kodak – an RPI alum invented the digital camera there in the 1970’s. Kodak dabbled with it but sort of squelched it to keep their film business alive. (“CMOS imagers could never replace film” was the cry). Not a perfect analogy, but Joseph Schumpeter’s 1940’s vision of “creative destruction” as the driver of capitalism is alive and at work in all of the lighting, LED, sensing and IoT churn we are seeing now. The dust might settle a bit in about a decade or so. Meanwhile, the luminaire industry may or may not profit from some of the things that are coming, but lighting designers, installers and the rest will have to know about them in order to deliver high quality of light and increasingly complex services related, in some way, to the luminaire.
From JOHN:
Thanks, BILL.
The reason I don’t agree so much is that I think luminaires are already componentized, power supplies have always been separate. I think you underrate the value of the metal benders outside of the Category B office world – there are trends, styles, change, which the tech industry isn’t ready for – nor the SKU problem – a single Metalux product can be eight choices of color temp, output, distribution, etc, totaling tens of thousands of incremental options. A premium project solution might be hundreds of thousands of options. Tech will cater to the power supply options, and possibly other bolt-ons. There will always be luminaire manufacturers.
What will be interesting is what happens as costs are reduced to the minimum production cost on products (as in, where fluorescents have been for a long time, metal body, bases, wires, ballast, lamp), then we’re back to the same place we were in the 1990s for commercial products. No margin for quality thinking, just basic CAT2 office luminaires or whatever. Then there will be a new generation of trends and we’ll start again.
From BILL:
Thanks, JOHN.
I did not mean to use “metal benders” as a pejorative term – and I do understand that just managing SKU’s is a challenge for most design projects. I will be talking to the Eaton CTO in a few weeks to discuss IoT and its role in lighting fixtures and future business development.
I agree that there will always be luminaire manufacturers, and increasingly, they will either pull IoT and smarts (with sensors) into their designs (part of their componentized process) or cede that to the networking and BMS guys and let them get the revenue from value added features and services, which with emerging trends in “wellness” and improved smart building drivers, will come.
I do very much appreciate your grounding me in the realities of the lighting design business – we should revisit these conversations in 5 years – likely for a good laugh!
