When you host a conference session about the Internet of Things but refer to it, for some reason, as the Internet of Connected Products, you should be ready for someone to poke fun at you. Leave to an industry analyst sitting on a panel discussion in that session to do the job.
“IoCP, which obviously refers to the Internet of Carrier Pidgeons,” joked Josep Di Paolantonio, an independent researcher based in San Francisco who took part in the track as part of Salesforce’s recent Dreamforce 2014. He got some laughs, but then quickly turned the pun into a useful analogy about the IP-enabled sensors embedded in all kinds of everyday products.
“In a way it’s actually like (carrier pidgeons), where you use its eyes as an extension of our own eyes. The IoT is really about sensors, and the data coming out of those sensors and what we can do with that . . it can help us define customers, open up new business models.”
Some of those new business models include predictive maintenance, insights on customer behaviors that can improve existing or new services and even the ability to sell non-competitive data to other firms in the same industry. Analysts on the panel, which was moderated by Salesforce CTO for Customer Connections Charlie Isaacs, suggested CIOs may still be struggling with figuring out what approach, if any, to the IoT they should take.
“There is still a lot of ‘bright and shiny object’ syndrome,” said Rebecca Wetteman, vice-president of Boston-based Nucleus Research. “There’s no new model for ROI on IoT. It’s just that it wasn’t cost-effective to do it before. What we’re seeing now is that the technology is offering breadth and repeatability — a sensor connected to IP are becoming more valuable the more people it touches and greater the frequency of how often something happens. Nobody wanted to spend the money on a smart fridge. Now, the cost is being brought down.”
Wetteman was most interested in projects such as one in South Bend, Indiana, where firms have reduced flooding in basements by putting sensors on drains. This allows them to watch for the impact of rainfall and open additional valves as necessary to manage flood water. CIOs shouldn’t limit themselves to merely tracking data via the IoT, however.
“The value is not just uni-directional, where you’re sending information out to a device, but multi-directional, so that as things become more intelligent they can participate in new workflows,” she said.
Although some might assume that effective use of IoT will also involve complex big data analytics, Di Paolantonio said sensor networks are offering a return to computational statistics that could offer greater accuracy.
“You don’t have to rely on probabilistic models but can bring in precise data from everything from ATMs to jet engines to manufacturing systems and handheld drills,” he said. “It will reach the point where it will tell you that this particular, very precise item is going to fail in the next 24 hours.”
Kate Leggett, an analyst with Forrester Research, suggested CIOs look for inspiration in areas from the consumer space, including home automation products like the smart thermostat Nest that was acquired by Google. The value from such early IoT products may help show enterprises the way.
“Try something in your personal life,” she said, pointing out that in some ways, creating an effective use case for IoT will not be a multi-million dollar initiative. “It’s cheap hardware and software — very small, very lightweight.”
Wetteman agreed, adding that enterprise IT shops need to democratize the skill sets around the IoT so that it doesn’t seem like the domain of a small set of experts.
“It’s putting the blueprints out there so you don’t need to be a data scientist,” she said. “There’s a rising role in every organization that’s adopted the cloud where someone in marketing becomes the marketing tech or the finance tech. We need to put the tools in place to help them find those opportunities and deliver value.”
Over time, it may not be that difficult to find them, said Esteban Kolsky, founder and principal with with ThinkJar.
“Everything will be generating and using data,” he said. “As close to never you’ll find a situation where connecting things is a bad idea.”