AIP Podcast Episode 69 - Check Engine Light for HVAC Systems by Talos IoT

This episode’s guest, Michael Scelzi, CEO and President of Talos IoT, joins host Anne to explore how IoT and connected systems are transforming one of the world’s most essential yet overlooked industries: HVAC. Drawing on 30 years of experience in energy and building systems, Michael shares why many HVAC units fail prematurely and how Talos IoT is addressing the issue with predictive diagnostics, digital twins, and real-time data insights. He opens up about the challenges of building reliable hardware, the lessons learned from costly connectivity gaps, and the breakthroughs that now allow Talos to deliver 100% data retention and optimize building performance. Michael also explains how commercial real estate firms and manufacturers can reduce maintenance costs, extend equipment lifespans, and lower their carbon footprints with intelligent HVAC solutions. Tune in to learn how Talos IoT is bringing the built world into the cloud era—creating smarter, more sustainable, and more reliable environments for businesses and communities worldwide. 

Follow AIP Affiliate, Talos IoT
Website: https://www.talosiot.com/
LinkedIn: https://www.linkedin.com/company/talos-iot/
YouTube: https://www.youtube.com/@talosiot324/videos

Follow AI Partnerships
Website: https://www.aipartnershipscorp.com/
LinkedIn: https://www.linkedin.com/company/aipartnershipscorp/
X: https://twitter.com/AIPartnerships

The AIP Podcast is hosted by Anne Cheng, on behalf of the AI Partnerships, a Railtown company

Transcript

SPEAKER_00 0:00

Now, global warming and the rising deployment of solar energy is making HVAC and air conditioning more accessible now more than ever. But a whopping 75% of all HVAC systems are still, unfortunately, installed wrongly, leading to a premature failure or requiring a repair that is still, by and large, a superbly manual process. When an HVAC system has a fault, a large number of people are still involved in restoring the system to working order. And thus enter Taylor's IoT, which reduces the time to value and helping us all over the world keep cool. Literally. Welcome everybody and welcome back to yet another episode of the AIP podcast where we speak to CEOs on the bleeding edge of AI and emerging technologies. Once again, I'm your host and turning on behalf of the AI Partnerships Corps. And today, our guest, Michael Skilsey, is the founder of Telos IoT. And he has spent the last 30 years in the energy and HVAC industry solving for some of the most practical problems, bringing to light some of the inherent challenges of systems that so often we take for granted. Michael, welcome to the show and thank you for being here with us today.

SPEAKER_01 1:27

Thank you very much, Ann. Pleasure to be here and appreciate you reaching out to me.

SPEAKER_00 1:31

Thank you. Michael, you've been around the block and then some. Tell us your backstory that led you to creating Talos IoT.

SPEAKER_01 1:39

So I was an HVAC contractor for many, many years and realized that the industry really had not innovated in any way, shape, or form. Still, to this day, a person, an occupant of a building or a house, or any type of structure, has to call somebody if they're hot or cold or they think something has failed. So that value chain has not evolved. So using IoT and things like that, I can make that process much smoother. Usually it's 10 steps for somebody to answer an HVAC hot or cold call. We bring that down to three steps. So the advent of connectivity in what they call Industry 4.0, taking the built world and digitizing it in the cloud is where I started about five years ago and we built our own platform for that. Specifically to take the value chain from 10 steps to three steps.

SPEAKER_00 2:37

How wonderful, Michael. Tell us a little bit more about Telos IoT. I mean, air conditioning has been around for such a long time, and yet the system is still broken and indeed riddled with folks throwing the problem over the fence.

SPEAKER_01 3:01

But there's no real feedback loop, right? Other than a human saying I'm hot or cold. So what we did was we built a LoRaWAN, which is long-range wireless area network solution where we pick up specific sensor points on the HVAC, basically systolic diastolic pressure. You know, we pick up the real details of being able to troubleshoot an HVAC unit. We take that through a LoRa router, bring it to our cloud, where we run diagnostics on it and predict failure and make sure time schedules are accurate and things like that. So that process took quite a while. Um, quite an interesting process. So building hardware is a lot harder than you think. Building software is not as hard as building hardware, and connectivity is harder than I thought. So it was a long time coming, but we are we're at we're in a good position right now.

SPEAKER_00 3:55

That's amazing. But Michael, I recall during our pre-show call, you mentioned that a really, really small glitch in connectivity cost you really big bucks. Tell us more, and how can people in industry be aware of the problem of intermittent connectivity? And really, what can we do to secure our businesses in an age where connectivity is so often taken for granted?

SPEAKER_01 4:18

So the first one was we were doing a utility pilot program, and we had 100 HVAC units connected, because the utilities know that when an HBAC unit's sick, it uses a lot more power, 33 to 35% to be specific. So they were interested because we're all constrained on grid. Now we don't have enough electricity, so they were interested in getting the data and the details, and you have to have 100% data, you can't have 99.3%. So we went out and deployed our system, and we used LoRa and we use a lot of off-the-shelf products, and we found that one specific modem lost 0.7% of the data. Uh, and that was a utility-grade rebate program, so it would have been over a hundred million dollar deal for us. Uh, so it's a little frustrating to say the least. Um, we solved that. Always learn, fail forward. We solved that by taking our existing board and putting an SD card on it that has a 60-day ride-through. So as long as the HVAC has power, I'm still recording all that data. So I have 100% data retention, 100% data connectivity after any issues occur with the router or LoRa network. So it was a it was a tough one to learn, but I learned it and we fixed it.

SPEAKER_00 5:36

Yeah, I think it it really underpins the fact that converged and secure networks are really, really hard to solve for, with increasing complexity through additional solutions that you're currently integrating, like air filter systems. How do you think you would help commercial real estate owners with preventative maintenance on their properties through digital twins?

SPEAKER_01 6:00

So it's really a monumental shift. So if you're in a building now, there's most like an air conditioner heating unit that has a filter that has no feedback, nobody knows what it's doing. We change clean filters since COVID, monthly and quarterly, and based on engineering analysis, meaning the pressure before and after the filter, we know exactly when to change it and we can predict it. So for some of the buildings, we're going one year without changing the filter, and this is based on engineering, not guessing. The industry now still does it four times a year in commercial real estate. If you go on the internet, everybody tells you you should be changing it once a month in residential, and the engineering just doesn't support that. The filter manufacturers do, but the engineering does not support that. So in general, we're seeing about a 50% reduction in filter maintenance and preventative maintenance on HVAC units because we're doing it every second of every day. Uh number one, number two, on the uh as far as run times and things like that, we found that 77% of HVAC units run all night, even though they have a programmable thermostat. So it the ROI on that is significant. Not to mention they we're not running the equipment as long. We're not running the car as long, so it lasts longer.

SPEAKER_00 7:19

That's crazy, and that is so important. Michael, this conversation has been really fascinating, but that's really all the time that we have for a conversation today. But before we go, tell us what's on your horizon? What are you looking to dive deeper into embedding your systems with large corporates or real estate companies?

SPEAKER_01 7:42

So there's a couple markets that we're going after right now. Number one, we're gonna have the largest regressible database of actual and factual HVAC data in the field with the right components on there. For the commercial real estate firms, we're going directly to the big guys like the large management companies, Collier, CBRE, Cushman Wakefield, Jones Line LaSalle. So we're going directly to them as a value add. For existing customers that or existing manufacturers, we're going to get share with them actual and factual data on warranty. Because as we mentioned in the beginning of the podcast, a lot of these things aren't installed right. So, as from a manufacturing perspective, number one, from a brand awareness perspective, having a better quality out there sells more product, number one. Number two, we also reduce the warranty cost because we have actual and factual data right after the startup. So we can resolve the problem on day one rather than day 60. On day one, it might be a $10 problem, on day 60, it might be a $10,000 problem. So there's a real return on investment there for the big guys. So manufacturers are primary and also commercial real estate firms as they are looking to reduce carbon and also maintain buildings at an optimal temperature.

SPEAKER_00 8:56

That's brilliant. Well, Michael, I've thoroughly enjoyed this conversation with you. And once again, to all our listeners, don't you forget to like, share, and follow our show and even send this to someone who would be interested in the subject. Until next time, my name is An Ching, and my guest has been Michael Skelsey of Telos IoT, is signing off on behalf of the AI Partnerships Corps. Until next time, cheers.