To gather insights on the evolution of IoT to this point in 2017, we spoke to 19 executives who are familiar with the current state of the Internet of Things.
We asked them, "What are the most common problems preventing companies from realizing the benefits of IoT?" Here's what they told us:
- Not breaking opportunities into winnable parts. Applying a centralized architecture versus an edge play. There’s time sensitivity to the data. Companies are dropping data into a Hadoop cluster and holding on to everything because they have figured out what data is gold and what is garbage.
- Not getting started. Not putting a strategy in place. Trying and giving up. Sees Hadoop shut down because didn’t have the people with the skills to operate and the company did not see a return on their investment. Find the right use cases and crawl, walk, run. We will work with clients to identify three or four use cases, rank them on a scale of the potential value to the business and difficulty of accomplishing. Identify those use cases that will deliver the greatest bang for the buck.
- Companies pursue IoT because it’s a novelty rather than a strategic decision. Everyone should be required to answer four questions: 1) What do we need to know? 2) From whom? 3) How often? 4) Is it being pushed to me? It’s necessary to get data to run your business, you need to identify the data that’s needed to drive decisions about manufacturing, product development, etc.
- Personally identifiable information (PII) is an issue. Need to identify a politically correct way of communicating. Aggregate data and use what you learn to help make people’s lives simpler and easier. Learn that the same person that bought a $100 pair of jeans at one store then went and bought a $500 coat at another store. Next time, the store that sells the jeans suggests coats.
- Concern around the security framework in industrial and consumer. Not thinking through security strategies and practices. Failure to articulate the business value.
- Management perception due to the hype cycle. Expectations are not realistic. Large enterprises are opening the back-end storage to push content and objects around. It’s much easier to do now versus a few years ago when everything was Microsoft.
- Companies struggle the first time they implement IoT by trying to do everything. The more successful companies focus on creating a competitive advantage and differentiation. They’ll bring in third party providers to help refine their offering and to stay focused on their core business. Don’t underestimate the complexity of an IoT transformation.
- Having the right approach to security and making security paramount.
- Everyone is, or should be, concerned with security. Exposure of intellectual property and putting data at risk. Pharmaceutical and nuclear are especially concerned about opening themselves up for more risk.
- 1) Huge capital expenditure. Can’t buy large scale M2M solutions off the shelf. 2) Huge risk of security with the exposure of high attack surfaces. 3) Concern around the potential for regulation that may step in like the FAA did with drones. These are high-level risks and there’s also a lack of technical expertise.
- A lack of utility to what they are building. Save energy and improve security. Offer useful services with ROI very quickly. Lack of a consistent ecosystem with Google and Apple on the consumer side and Cisco and IBM on the industrial side.
- There is a wide variety of connected devices and solutions on the market today. But they do not talk to each other; they are not interoperable. They use different communication protocols. Instead of relying on well-known, reliable wireless communication protocols like Wi-Fi, ZigBee and Bluetooth, the world’s top technology companies are battling for market share, trying to be the only Smart Home Technology provider. Instead of working together to provide easy-to-use, interoperable services, and devices, they are creating proprietary ecosystems with protocols and “standards.” On the performance level, there is not much that truly separates these companies. Making a decision about which technology to use (Apple, Intel, Google, Samsung, etc.) is a daunting task even for sophisticated engineers. Consumers cannot possibly be expected to figure it out.
- Lack of use case and strategic planning. Healthcare is probably doing a better job than most. They will identify the use case and create a roadmap showing when the ROI will be realized.
- Data privacy laws. Proper data architecture. IoT will consist of anywhere between 20 billion and 50 billion sensors by 2020. How do you turn data from a million to a billion devices into meaningful insights? IoT devices are not being configured to speak a single language. Until a single communications standard emerges to organize the IoT around a shared language, BI systems are to provide the single, massive interconnected networked of analytics. This is the same problem that we have had with business applications and databases – each use a different API and different SQL dialect. The same applies to IoT – the data language for each manufacturer and for each device is different, so data normalization and networking of different analytic applications need to happen outside in the BI system. Commercialization. Beyond the data science and data engineering aspects of an IoT project, you need a team of people to realize the commercial value from your data. Similar to how you would create and launch a physical product, you should think about an IoT data product. The same principles of design, engineering, pricing, and go-to-market apply. One thing is easier, though: creating data products and monetizing them happens a lot faster as you leverage the cloud as your development and distribution platform.
What do you see hindering the realization of benefits of IoT?
And in case you're wondering, here’s who we talked to:
- Scott Hanson, Founder, and CTO, Ambiq Micro
- Adam Wray, CEO and Peter Coppola, SVP, Product Marketing, Basho
- Farnaz Erfan, Senior Director, Product Marketing, Birst
- Shahin Pirooz, CTO, Data Endure
- Anders Wallgren, CTO, Electric Cloud
- Eric Free, SVP Strategic Growth, Flexera
- Brad Bush, Partner, Fortium Partners
- Marisa Sires Wang, Vice President of Product, Gigya
- Tony Paine, Kepware Platform President at PTC, Kepware
- Eric Mizell, Vice President Global Engineering, Kinetica
- Crystal Valentine, PhD, V.P. Technology Strategy and Jack Norris, S.V.P., Database Strategy and Applications, MapR
- Pratibha Salwan, S.V.P. Digital Services Americas, NIIT Technologies
- Guy Yehaiv, CEO, Profitect
- Cees Links, general manager Wireless Connectivity, Qorvo
- Paul Turner, CMO, Scality
- Harsh Upreti, Product Marketing Manager, API, SmartBear
- Rajeev Kozhikkuttuthodi, Vice President of Product Management, TIBCO