On Friday, March 13, at SXSW 2015, representatives from Ayla Networks, Echelon Corporation and Xicato will join moderator Mike Krell of Moor Insights & Strategy for a lively, wide-ranging panel discussion on smart infrastructure, the Industrial Internet of Things (IIoT) and some rules of the game that will spell success in this high-profile market. The panel, titled Network All the Things: What About Infrastructure?, will take place from 12:30 – 1:30 p.m. in Room 208 of the JW Marriott, 110 E 2nd St., Austin.


Panelists Dave Friedman, CEO and co-founder of Ayla; Ron Sege, CEO and chairman of Echelon; Menko Deroos, CEO and co-founder of Xicato; and Mike Krell, analyst, Internet of Things, for Moor Insights & Strategy, sat down to answer some questions that will preview topics the panel is likely to cover. Given both the areas of overlap and the points of divergence on major themes, this panel promises to be illuminating.

Mike, as the moderator of the panel, what do you see as the main themes you’d like to see emerge from the discussion?

MIKE KRELL: The first is that while most people think of the IoT as something that will happen ‘tomorrow,’ in fact, it already exists, especially as the Industrial IoT. It’s not a brand new market—it’s an expansion of traditional machine-to-machine (M2M) markets—but it does have a lot of cool new things within it.

Then, there are things like the problem of security as more things get interconnected, and the role of standardization in propelling business models and revenues in the IoT. Also, I’m hoping the panel will explore whether ‘brownfield’ environments, which we usually think of as staid and boring, actually offer greater potential for money-making than ‘greenfield’ environments, which are more traditionally associated with innovation.

OK, with that overview in mind, let’s ask the panelists: What’s your view of ‘smart infrastructure’ as it relates to the IoT?

RON SEGE: Smart infrastructure is to the IoT what computer networking was to the World Wide Web. If there’s no optimized way to connect devices to other devices and to the cloud, there’s no IoT. Silicon Valley is fixated on the IoT meaning big data and the cloud, and they just assume you can get the data from all the ‘things’ involved. But reliably collecting and propagating that small data into big data, to ultimately take action, takes lots of innovation and smart technology—a smart infrastructure. 

DAVE FRIEDMAN: We think in terms of configurable infrastructure. If you’re building IoT products, you need to be able to view the IoT as infrastructure that can power your business. Just like you’d use Salesforce or Oracle or SAP, you need to have available all the elements of IoT infrastructure that you can just use without putting it all together yourself. When I say ‘infrastructure’ for the IoT, I’m not talking about a gateway, or a Wi-Fi connection, or just a cloud, and it’s not applications. It all has to seamlessly work together as a complete entity, and it has to be configurable to suit the needs of any particular business.


MENKO DEROOS: Xicato is a smart lighting company focused on retail and hospitality markets. So for us, smart infrastructure is the infrastructure around lighting. And instead of talking about the Internet of Things, we talk about the Internet of Lights. In the Internet of Lights all the lighting components are getting wired and wirelessly connected. And all these lights, sensors and controls get configured through applications and cloud-based services to implement rules on all those components, which makes the lighting infrastructure to become smart.

What do you see as the major challenges today for the IoT?

MENKO DEROOS: The problem is not building the actual infrastructure in the Internet of Lights—because we see an opportunity to use smart lights as the infrastructure for almost no additional cost. We envision the lights becoming the gateways and routers in the Internet of Lights. The problem for the industry is to agree on providing standardized open APIs to set up open ways of communicating with everything that’s in there, so that the things of the Internet of Things can connect to those lights—which in the lighting world is not the way of working today.

RON SEGE: In the Industrial IoT in particular, a huge challenge is that there are already a billion devices connected using existing technologies. These legacy devices need to evolve to converge on the IoT, without ripping and replacing the devices. Establishing a migration path is not easy, but it’s necessary for many IIoT situations. In addition, the use cases for these existing devices require unprecedented reliability, scalability and security—beyond what computer-networking technologies can provide today.

DAVE FRIEDMAN: There are literally billions of different system architectures that are part of the IoT. A lot of people in the technology space who talk about the IoT have to do a lot of hand waving about how you manage the massive fragmentation that is characterized by the IoT. Many of today’s connected devices, such as laptops and mobile phones, have complex architectures with full-blown operating systems. But many of the ‘things’ of the IoT—like a coffee pot, for instance—have tiny little microcontrollers and nothing like an operating systems, and they’re doing all sorts of different things, and somehow it all has to work. There’s a huge challenge in solving how everything gets connected and stays connected and just works.

Where do some of the greatest opportunities lie?

DAVE FRIEDMAN: The IoT, just like the Internet itself, is all about collecting and being able to manage and make sense of data, and then eventually to make better decision based on the data. For example, manufacturers that establish feedback loops because they have connected the ‘things’ in their manufacturing plant will win over their competitors because they’ll be able to change their business processes around maintenance and warranty and service—and even understand better what features to deliver in their end products—based on this feedback loop.

MENKO DEROOS: We don’t see markets in terms of brownfield vs. greenfield opportunities, but more as above-the-ceiling or below-the-ceiling opportunities. Above the ceiling it’s difficult and expensive to make structural changes. Below the ceiling is much easier and cheaper. And that interface of the ceiling is exactly where the lights are, and where we believe the opportunities are. The markets we look at are residential versus non-residential: How many hours a day are the lights on? How much energy can be saved by smart lights? But also: How can smart lights improve sales in a retail store?

RON SEGE: Although the consumer IoT gets more press attention, the real financial and societal potential lies in the industrial IoT. The cost of connecting to the IoT is dropping. Ten years ago, it was about $10 per connection, now it’s $4 for a Wi-Fi module or Echelon chip, and soon it’ll be $2. Also, and this is important, customers don’t really buy an ‘IoT.’ They buy solutions to certain vertical market problems. As for vertical market opportunities, we see lighting as the one with the biggest immediate potential, because connecting highly controllable and sensor-enabled LED lights can have many, many benefits in terms of health, customer satisfaction, asset utilization and learning. Beyond lighting, we foresee big opportunities in transportation—such as locomotion, aircraft telemetry and smart traffic systems—and the smart grid.

What would you say are some misconceptions about the IoT? And does the IoT or smart infrastructure have any ‘dirty little secrets’?

MENKO DEROOS: The biggest misconception is that the IoT requires lots more innovation before it will be real. Everything we talk about today, from a technology perspective, is available now. All the ‘things’ can’t be connected using Cat 6 cables, they don’t all have enough power to connect to routers over the long range needed for wireless connections, and it’s not always cost-effective to give them an IP address. But all the things are close to lights, which with their integrated power can detect those ‘things’ in their proximity and act as gateways to central services. You don’t even need to control or manage smart lights and connected things centrally. There’s control technology that uses swarm or whisper modes to communicate to things in the same way flocks of flying birds communicate.

RON SEGE: The dirty little secret is that there is no IoT. What we call the IoT, or smart infrastructure, is a confluence of technology factors that allows certain vertical markets to benefit from the combination of low-cost connectivity, cloud management and big data analytics. At the end of the day, the ‘IoT’ will be a collection of applications that have real economic and societal value – enhanced productivity, better learning, better health care and so on.

DAVE FRIEDMAN: The misconceptions are that it must be easy to do, or that you can or must provide the whole solution yourself. When you’re working with giant systems like a computer or a phone, you have a huge amount of resources to work with. That’s not the case with the IoT. When you hear about security holes in IoT, it’s not because SSL somehow stopped working. It’s because someone had to cut corners because there weren’t enough resources on the ‘thing’ they were trying to connect, to manage and provide appropriate levels of security. It’s really hard to make a light, or a motion detector, or a door lock as secure as your computer. In addition, it’s not true that if you know how to build air conditioners you’ll clearly know what an Internet-connected air conditioner will do.

How do you see the IoT evolving in the future?

DAVE FRIEDMAN: It’s not going to be one size fits all. It will be trusted clouds talking to each other. There will be industrial routers that makers of venting systems, for instance, can connect and use as gateways to their trusted cloud. The applications will come together to combine the data from the trusted clouds to make broader sense of the universe. But flexibility will be key. One of the most critical features will be the ability to future-proof your products, to be extremely flexible with how they evolve. That goes all the way down to the PHY layer. No one has a crystal ball good enough to really predict the future, so you need to be able to change and reconfigure things flexibly.

RON SEGE: The IoT is already being called an over-hyped technology, and not everything being touted as IoT will be successful. There will be more attention on figuring out the business and societal benefit of what you’re bringing to market for the IoT. The ones with real business benefits will survive and thrive.

MENKO DEROOS: People in different industries, such as lighting, will work more closely with others outside our own industry. Lighting needs IT, and IT needs to consider the perspectives of lighting. Also, there will be a paradigm shift in how we build buildings, driven by the need to integrate lighting and other systems together, and enabled by LED technology and the Internet of Lights.

Any final words of wisdom to impart?

RON SEGE: One interesting tidbit is that the whole concept of smart infrastructure or the IoT is turning Moore’s Law on its head. Moore’s Law, which arose in the semiconductor world, says that every 18 months or so, for a given processor price, you’ll get twice the performance. In the IoT, every 18 months or so, for a given performance level, the price is cut in half. It’ll be interesting to watch how that plays out over time.

DAVE FRIEDMAN: Although aggregated data is very interesting in the IIoT, there’s only one company in the world that can really do much with the unique data to any particular device and that’s the guy who made it. Each manufacturer will need to get access to its own data and have warranty, maintenance, service and predictive analytics specific to its own products.

MENKO DEROOS: The three most important things to remember about the IoT are the importance of security; the need to think in terms of solutions, rather than components; and the need for open APIs.