AWS VP of Internet of Things shares his thoughts on avoiding the buzz, and the importance of embarking on an IoT journey only once a problem has been defined.
There are varying estimations on how many billions of devices will be connected to the internet in the near future, but what is certainly plausible is that the number of Internet of Things devices will well outstrip the number of people using them.
Enterprises, however, must be cognisant of not flooding the market — and in particular their business — with the unnecessary simply because of a fear of missing out.
According to Amazon Web Services VP of IoT Dirk Didascalou, it is important to understand why an organisation wants to do IoT and not just slap sensors onto things because it’s the current buzz.
“Sometimes we are a little buzzword driven — IoT will solve all my problems, AI will solve all my problems, now we have the new buzzword, AIoT — that’s why we typically start all of our conversations with a question: If you knew the state of all of your assets, all of your things, everything, in your company and you could reason on top of that knowledge, what problems would you solve and why?,” he told ZDNet during AWS re:Invent in Las Vegas.
“We ask about problem statements: What’s hard for you, where do you lose money, what is really a headache for you? We look at dedicated problems and once we understand what are really big business problems for our customers, then we work backward and figure out what can we do with IoT technology to solve them.”
Didascalou believes this is the only way IoT really brings business value, and said that the biggest challenge for companies is to figure out if they even need to add sensor technology, pointing to a trivial concept such as putting a sensor on a water bottle to measure how much liquid is left.
“One of the hard learnings that we had to go through … many of those [proof of concepts] never made it to production and the reason is, you never had a real reason, a business problem,” he explained.
“Go to the factory and ask, ‘Where do you struggle every day?’ and figure out what the product team doesn’t know about their product, or why the customer is unhappy and work backwards.”
He said IoT in principle addresses two major areas: Organisations can now create products and services which were impossible before; and also gives them the opportunity to something they did before, but better.
“The moment you know what your problem is, that’s already half of the success,” he said.
Didascalou pointed to UK-based company Vantage Power and its CEO Alexander Schey’s mission to reduce the number of people that end up with respiratory disease.
Vantage Power also electrified all of the red buses in London, but the company hit a roadblock and had to break down its problem into smaller parts.
“Breaking a bigger problem down, the company put these new electric motors in these buses and they all worked great in the beginning and then they started failing. That’s not a good idea if you’ve got a full bus of people and it doesn’t work,” Didascalou explained.
“Because it’s not yet matured technology, we need to figure out how they work, and what they did was they were ingesting all of the measurement data from all of the batteries — one trillion data points — they used analytics to process in SageMaker, they built a model using AWS IoT Greengrass.
“It’s mind-blowing, they can predict a month in advance if one small cell in any of their batteries will fail, and then they can arrange a service.”
Date: January 16, 2019