IoT devices are entering the workplace in all shapes and sizes, from workers wearing smartwatches to industrial sensors such as soil monitors. The data pouring in may be so overwhelming it’s unclear what should be done with it, why, and what the risks might be. Here are a few ways to navigate the maze.
As wireless sensors find their way into a wide array of products – both for consumer and industrial purposes – IT organizations of all types are grappling with what the internet of things (IoT) means to their operation. According to Gartner, 20.8 billion connected “things” will be in use worldwide by 2020.
Unlike some other technologies, IoT tends to have a multidimensional effect which not all companies are equipped to handle. IT leaders need to consider how IoT will influence the company’s business strategy and its business processes, as well as potential effect on partner and customer relationships. They also have some very real IT issues to consider, such as infrastructure.
“Organizations are not planning for the infrastructure they will need to store and process all of the data that will be generated by IoT devices, and they are not properly securing information as it is being transmitted from these devices or even as the data sits at rest,” said Tim Tutt, CTO of IT management and engineering solutions vendor Bogart Associates, in an interview. “Not planning for the infrastructure need can result in all types of system failures, while not having a strong security plan around the data can expose an organization or its customers to malicious actors.”
Adding to the potential confusion are differing perspectives on even the most basic elements, such as how long data should be kept, where, and why.
“The biggest mistake companies make is when they implement IoT without determining a clear business case first. IoT is seen as a ‘cure all’ and companies need to be focused on the business problem: monetization or operational efficiency,” said Debbie Krupitzer, IoT North American Practice lead at consulting, technology, and outsourcing services company Capgemini, in an interview. “IoT is seen as a solution, as opposed to an enabler, for business outcomes.”
Mail and messaging management equipment provider Pitney Bowes has gone to considerable lengths to ensure its IoT efforts are successful. According to Pitney Bowes SVP of Technology James Fairweather, the company created a comprehensive employee training program and formed a small, central team to own Pitney Bowes’ Big Data platform. It also worked with partners to provide the data connectivity and transport layer and to stand up the data lake infrastructure. Its own platform is ingesting and transforming data from a multitude of existing systems into its data lake, so the data can be leveraged in a common data platform alongside new, IoT-sourced data.
“Companies are wrestling with the same challenge, which is how to succeed in both physical and digital worlds,” said Fairweather, in an interview. “Implementing these practices has allowed us to enjoy significant success with IoT.”
Other organizations aren’t as fortunate. A number of obstacles can get in the way, some of which are internal and some of which are external. We’ve identified some of challenges and best practices. What’s your experience? Tell us all about it in the comments section below.
Think Differently
The old way of thinking may not serve a company well as it ventures down the IoT path. Organizations considering IoT to be simply another new technology could be missing the big picture view of how IoT will change the business.
“[It’s a mistake for] organizations [to] treat IoT as simply an enhancement to existing products, rather than looking afresh at whether a product or service could be completely redesigned, reengineered, or delivered in a completely different fashion now that IoT information is available,” said William Webb, IEEE Fellow and CEO of nonprofit IoT standards organization Weightless SIG,in an interview. “For example, considering a connected car as a four-wheeled vehicle which can now be updated remotely, rather than thinking of it as a travelling computing and entertainment platform.”
Decide How The Data Will Be Used
While it’s not possible to anticipate all the ways IoT data might be used in the future, it is important for organizations to put some thought into how they intend to use their IoT data and how it aligns with their business capabilities.
“As with any other enterprise architectural decision, considerations will have to be made pertaining to the persistence and disposability of the data and insights related to its use. Organizations will also need to determine if the IoT data is for automated processes or enterprise analytics which will cause decisions of service-oriented architectures, in-memory systems, or persistent data lakes,” said Michele Goetz, VP of Corporate and Product Marketing at self-service data preparation platform provider Paxata, in an interview. “Big Data architectures will need to incorporate Spark and Kafka to more easily handle streams and processing.”
Focus On Business Value
IoT can have a transformational effect on a business. Some traditional watchmakers, such as Tag Heuer, are a good example of this. They’re not only making digital versions of analog watches anymore.
“The true value of the Internet of Things is that [it] potentially provides vastly superior tools for probing and understanding the world of humans. But if current [business] behavior around existing Big Data opportunities is any indication, a huge number of them will be content to remain in their ossified data and IT infrastructures, looking at colorful Excel charts, telling them the counts of things they already knew to ask about, instead of predicting the future along dimensions no human had ever thought of,” said Peter Wang, CTO and co-founder of the Anaconda open data science platform creator Continuum Analytics, in an interview.
Think Broadly
It’s easy to look at IoT as an option to solve for a very particular business goal, such as reducing the number of times a field service company has to send out technicians. By looking too narrowly at what IoT can do, potential business opportunities can easily be overlooked.
“CIOs are increasingly being tasked with owning the responsibility for these IoT initiatives due to the cost-saving and revenue generation potential, particularly when you add analytical platforms to derive insight from the sensor data volumes and integrate those systems with ERP, billing, [and] warranty management systems to automate business processes,” said Jeff Healey, director of product marketing, Big Data Platforms at enterprise technology solution provider HPE, in an interview. “Working in silos is the biggest mistake that companies are making, [for example] only focusing on reducing ‘truck rolls,’ rather than taking a concerted and strategic approach to how to capitalize on this opportunity.”
Assess “Value” Carefully
Companies need to consider not only the ways IoT devices could benefit their organization and its customers, but also how the data could be used by third parties.
“It’s important not to underestimate the value of even the smallest piece of information. For example, data from a traditional motion sensor that is used to detect burglaries and trigger alarms can now be used in many different ways,” said Kevin Meagher, SVP of business development at smart home product vendor Roc Connect, in an interview. “It can show that a property is occupied, which could help insurance companies assess risk since occupied properties have less risk. The same detector and data can be extremely valuable in a healthcare application. For example, the elderly living on their own.”
Contemplate Data Retention
There are different schools of thought on deciding which data to keep and which to jettison.
“Most data produced by IoT are sensor/actuator data, inherently valuable for real-time processing. A few seconds later might be too late when you’re monitoring a sensor controlling a 12-ton crane, but acceptable if you are monitoring the daily rain on a farm,” said Evaldo de Oliveira, director of business development at multimodel No+SQL database technology vendor FairCom in an interview. “Governance is a key factor here, and storing data for auditing purposes is even more critical. In most cases, I think storing sensor data is not necessary past a few days for operations technology, but is probably very relevant for legal departments.”
Decide Where Analytics Will Take Place
Which data should be analyzed at the edge and which data should be analyzed centrally? Should some analysis be done in between the core and the edge? The choice may depend on many things, such as the company, use cases, potential value, potential risks, sensitivity to latency, the data itself, and more. Sometimes, all the moving parts cloud thinking.
“Insights and analysis that can be completed without too much history and without much context from other devices should be analyzed and acted on at the edge. Anything that requires historical correlation or correlation across zones and sets of devices [is] best analyzed closer to the core,” said Goutham Belliappa, Big Data and Analytics practice leader at Capgemini, in an interview. “In a mature, complex IoT ecosystem, we see a truly federated approach, where some analysis and action occur at the edge — often in stream — some analysis and recommendations occur in interim lakes between the edge and the core, and some action that requires extensive correlation and history occurs at the central core.”
Consider The Risks
One school of thought is to save all IoT data, at least until its value or lack of value can be determined. After all, storage, processing, and bandwidth costs continue to fall and data lakes can scale.
“While data scientists will see archives of IoT traffic as hidden assets, a more cynical worldview would be to question the wisdom of creating vast lakes of data whose value is asymmetric and that you don’t have a business use for,” said David Rolfe, director of Solutions Engineering EMEA at in-memory operational database vendor VoltDB, in an interview. “Unguarded data lakes represent a serious security risk and may, in fact, be the potential source of lawsuits.”
Think Like A Hacker
IT organizations grappling with IoT need to consider how it fits into, and potentially affects, existing IT infrastructure and security. Otherwise, the benefits of IoT devices may be outweighed by their risks.
“The Internet of Things is fraught with peril. Every new device that becomes wireless-enabled is now a new device that can be hacked. Whether it be Target’s HVAC system or the recent Volkswagen key hack, multiple sets of radios have to be considered,” said Hyoun Park, chief research officer at research and advisory services firm Blue Hill Research, in an interview.
One of the differences between mobile security and IoT security is that there are far more radio frequencies and signal standards that may need to be considered. Which security standards are relevant depends on the hardware and signals put in place, Park said.
Don’t Downplay Security Breach Potential
Organizations should take a proactive approach to IoT security to minimize the risks of potential data theft, legal exposure, damage to reputation, and danger. While IoT devices are not new, the technology category as a whole lacks some of the safeguards that have been established for more mature technologies.
“Many organizations haven’t even wrapped their heads around mobile devices and are still struggling to deal with the Bring Your Own Device phenomenon. As a result, IoT devices are not even considered when it comes to enterprise security models,” said Srinivas Mukkamala, CEO of cyber risk management company RiskSense. “The challenge organizations face is that IoT devices not only lack a common security framework, but more importantly they are not yet covered by traditional security tools. This makes it difficult to monitor them for threats and mitigate problems.”
Unlike traditional cyberattacks, IoT incidents are not limited to extracting information. Instead, they can be used to cause physical harm and exploited by state-sponsored cyber attackers to wreak havoc, Mukkamala said.
Beware Of Potential Correlations
IoT data being used for a legitimate business purpose could, nonetheless, compromise private information.
“Companies often don’t realize how personal the data a connected device might be collecting really is. What may seem like lots of temperature readings on a connected thermostat could potentially reveal when people are home, for example. If this data is combined with [personally identifiable information], it can be both extremely valuable and extremely dangerous,” said Calum Barnes, senior manager, IoT Products and Solutions at IoT platform provider Xively by LogMeIn, in an interview. “Companies should focus not just on protecting data from data breaches but also ensure their terms and agreements and privacy policies are in line with how the data will be used.”