Is Power Over Ethernet all it is cracked up to be?

22 November 2023

I think one of the first Smart Building events I attended was at Cisco’s Bedfont Lakes HQ. It wasn’t that long ago and I remember the buildings there being more sparsely occupied compared to those shown in presentations by the hosts. Since then I have seen a fair few more from them including at events I have curated (see here and here), at Smart Buildings Show and elsewhere. Personally, I have found them way too salesy and to paraphrase:

It’s not about technology, it’s about people. Hey look at our new technology.

I guess that technology muscle memory runs deep and there are quarterly tagets to meet. But as former colleague at one of their vendor partners pointed out, what Cisco are good at is templating their technology from one domain and wrapping a seemingly coherent story around it when trying to make inroads into another.

In the smart buildings one that story includes some of their own buildings (mostly in USA) and one or two stand out examples like 22 Bishopsgate where detractors would say they overstate the importance of their technologies in the deployment of smart assets. And particularly when it comes to the role Power Over Internet (PoE) plays in that story, as it gets hammered home with slides on the reduction of energy waste by PoE’s use of DC power, the resource time saving of installing it, reduced cabling costs and more depending on duration of presentation.

More recently that PoE story has been incorporated into the marketing of their Catalyst switch product range with the bold claim, “When Switchports become Smartports, Buildings become Smart Buildings….” (see here). That seems quite a stretch and so it has been interesting to probe a little deeper into the POE story. And not least because when I have asked both (smart) buildings consultants who specify technologies and those at the (master) systems integration sharp end of deployment. The usual response is that it is not surprising that Cisco make a big deal about PoE as it helps them sell a lot of switches.

That’s usually followed by them repeating “a lot” but often with a more detailed explanation about the following:

  • Geographical location
  • Technological direction of travel
  • Types of project

Geographical location:
As an aside, North America may not be benchmark of what’s possible in smart buildings having spoken to building consultants with global operations. That’s probably one for a different discussion but relevant if you are using examples from there as part of what can seem like a pitch. But coming back to North America there maybe a ‘labor’ cost equation when it comes to AC installation compared to DC that is peculiar to that market. I am not sure if that is about regulation on who can install which and/or if that is a electrical contractor union issue linked to lower voltage and safer structured cabling vs electrical conduit. There are also other parts of the world that those I have spoken to where PoE being DC has some significant safety benefits given lack of electrical regulation or at least enforcement if it does exist. Point being that installing PoE can make financial and safety sense depending on the geographical location of the project. And it’s probably worth mentioning that I have installed Cat 6 cabling in my home myself to connect up WiFi access points.

Technological direction of travel
I have heard smart building specifiers, MEP consultants who run projects and Master Systems Integrators (MSIS) all wax lyrically about MQTT and PoE enablement of edge technologies, along with how likes of Delta Controls are leading the way. I am not going to go into any depth because that is definitely one for a different post, bur it would seem to be another plus for the PoE deployment case.

Types of project
But I also hear from those just mentioned that the pros of PoE are somewhat dependent on the project scenarios, i.e. New/rebuilds, refurbishment or retrofit. Specification in the first two of those is often starting from a blank piece of paper, particularly if the refurbishment is major one where the building has been stripped back to the shell and core. Although in reality that blank piece of paper is far too often a specification copied and pasted from a previous project. There are those way better qualified than me to advise on the merits of PoE deployment on those projects and not least because of the direction of technological travel mentioned above. The point being that new/rebuilds and (major) refurbishments could also be seens as a yay, hay, hooray for PoE depending on who you speaking to and their architectural philosophy although that isn’t always neutral.

But what about retrofits?
The perfect storm of higher interest rates, back to office challenges and sustainability matters I wrote about recently for Smart Buildings Magazine (see here) means that in most geographical locations it’s the retrofitting of existing (physical) assets that is top of minds for their owners and landlords. And with the retrofitting of smart buildngs, the jury on PoE is out because there’s often a lot of existing technology assets that can be repurposed not least being lighting systems running on DC that run throughout a building like a spine that sensors can be connected to and are being, i.e. so smart building platforms can get data and analyse it. At the same time, there are likes of LoRaWan technologies for connecting with IoT devices like sensors, as well as ambient light sensitive photovoltaics to help power them.

There’s longer list of technological innovation, but 2 points linked to what’s mentioned in the paragraph above. Firstly, retrofitting offers an opportunity to be more agile than new/rebuilds and by using technologies as part (rapid) prototyping that can be quickly deployed in order to get data to help inform decisions about the technology upgrade road map. And the starting point for that not being PoE deployment and (converged) network architecture refresh. That’s not to say those won’t be part of the upgrade roadmap later.

At the same time, retrofitting is usually framed in terms of the repurposing of existing physical assets, and that being part of the rebuild vs retrofit cost calculation including the factoring in of sustainability considerations. As I plan to mention in a forthcoming article, what’s not clear is whether those calculations accurately reflect how much existing technology can be repurposed with upgrades and better integration thanks to ongoing edge innovation in smart buildings. And because that’s rarely the purview of those making those cost comparisons. I mention this because replacing what is already there may not be more sustainable than a PoE installation and network upgrade.

Ultimately, it doesn’t matter to me either way. This post is in part about the difference between a product sales pitch and conversation about how smart enablement is being done in what scenarios and trying to unpack the why of that. As mentioned in my, Unified Name Space a single source of truth? post, my reason for doing so is as follows:

I’m interested to know if and how there could be some kind of minimum viable technological foundation for smart enablement that could be built upon and scaled, particularly in retrofit scenarios. And if there is some consensus about the architecture/design philosophy of that. This would need to include the commercial realities including cost of technology. But also those linked to any potential skills gap when it comes to deployment and management of that foundation or whatever best describes a ‘modern distributed architecture.’ Maybe that’s one for another round table.

Finding someone to help fund my time to faciliate a round table like that or research more broadly has, sadly, proved more challenging. But I was glad to find I wasn’t talking complete nonesense having read the following from IIoT & Industry 4.0 Solutions Architect & Online Educator Walker Reynolds in his report about my post (see more here):

The short answer is, the market isn’t ready for standard foundation but it’s getting very close, but that doesn’t mean the foundation doesn’t exist — 2024 is the year the conversation will take form and I suspect that conversation is going to start at Connack in Munich next month. I will be going with ideas collected from the Industry 4 community to prime the MQTT community for this discussion on standards — which is the last domino to fall before we can take what has been developed through community innovation and codify it for scale. That being said, there most definitely is a foundation to build on and is being leveraged at scale in orgs on the upper half of the Industry 4 maturity distribution, developed across verticles and industries, hardened by the Industry 4 community.

Would be great to see that demo he mentioned and possibly make that a webinar so I can help amplify. I would also be interested to hear Walker’s thoughts on PoE and others I haven’t spoken to about the topic yet. Likewise, I would be interesting to hear from team at Cisco about if and how they see likes of Unified Name Space being part of their (smart building) architectural design thinking/philosophy. In the meantime, check out this post about the Real Estate Reimagined: better tech, smarter buildings web series from their Hybrid Work Center of Excellence team. It’s probably the latest you can see on where their smart building thinking is at with the second episode on Converged Network Architecture today at 1pm EST (see more here).

About Me

I’m an author and market engagement specialist who helps joins the dots between research, innovation, go-to-market activation and business transformation around 3 core activities: content, connections and conversations. Connect with me on LinkedIn or get in touch there.