Underwriters Laboratories (UL), the 1894-founded certification and compliance company, has truly materialized the testing needs of the internet of things (IoT) era. The IoT – well, the consumer IoT at least – is about the interconnection of computing devices within everyday household objects. Since that’s the case, it would make sense that a strong testing ground for it would be a house.
Enter the UL “Living Lab.” The lab is a two-story residence that provides a space in which devices can interact within a real-world setting so that these environments can operate quickly and coherently, without security compromises or interoperability snags. At the Living Lab, people within the house use various IoT devices to verify that they function in accordance with one another and the external world.
A few of the factors that are of greatest concern to the UL researchers within this environment are typical ones that impact network and device performance:
- Floor plan – how ceilings and walls might interfere with connection
- Noise – the influence of ambient noise from residents or other “things”
- Acoustic elements – the impact of furniture, drapes, rugs, and carpets
- Other Wi-Fi – additional radiating devices, including nearby Wi-Fi networks, that interrupt your own system’s communications
- IoT overload – bandwidth consumed by many different devices.
Essentially, this project by Living Lab is allowing them to uncover issues in a sort of “fishbowl” setting. From a more general perspective, the challenges of IoT technology can be understood through a framework provided by Ahmed Banafa of the University of California, Berkeley. His lenses through which technology of the IoT can be understood are security; connection; sustainability and compatibility; standards; and derivation of insights for intelligent action.
Security is a central concern of the internet of things. With all the new nodes come new ways for hackers to find their ways into the network – especially since devices are often not built with strong security in mind (because the IoT is growing so rapidly now, with a focus placed more substantially on function than on data protection).
How critical is security to the IoT? Look no further than this November 8, 2017, headline by Charlie Osborne of ZDNet: “IoT devices are an enterprise security time bomb.” The evidence comes from Forrester Consulting. The analyst’s poll of 603 line-of-business and IT executives at large companies from six nations (including the US and UK) found that 82% of respondents said they would not necessarily be able to pass an audit since they were incapable of locating the devices on their networks that were either operational technology (OT) or internet of things.
Partially due to this lack of knowledge related to the technology, the stresses of the internet of thing are real as well, according to the survey. 54% of people reported that the IoT is a cause of stress: they feel unsure that it has the protection they need.
Curiously enough, the ZDNet piece reveals one of the problems holding back security: unsureness about the IoT itself. Companies typically were not investing large amounts in internet of things projects, in part because executives were still rather reserved on the topic. With tight budgets, 2 out of 5 staff members polled said that their organizations were using traditional tools to protect IoT systems.
“This is a glaring issue for today’s firms, which need crystal-clear visibility into networks where BYOD and IoT are common,” said Osborne.
Connectivity is another basic concern of the IoT that will push us beyond the server/client communication paradigm that we have used previously for node authorization and connection.
Server/client is a model that is well-suited to smaller numbers of devices. With the advent of the IoT, though, networks could require the integration of billions of devices, leading to bottlenecks in server/client scenarios. The new systems will be sophisticated cloud settings capable of sending and receiving massive amounts of information, scaling as needed.
“The future of IoT will very much depend on decentralizing IoT networks,” noted Banafa.
One way that decentralization is achieved is by transitioning certain tasks to the edge of the network, as with fog computing architectures that use hubs for mission-critical processing, with data collection and analytics through the cloud.
Sustainability / Compatibility
Currently, there are many different companies using various protocols that are trying to develop the standards for the internet of things. This free-market competition can boost innovation and options; however, additional software and hardware may be necessary in order to interconnect devices.
Disparities between operating systems, firmware, and machine to machine (M2M) protocols could all cause challenges in the IoT.
The reason that these two elements, sustainability, and compatibility, are discussed under the same heading is that the notion of compatibility is directly linked to the ability for a general ecosystem to survive long-term. Some technologies will inevitably become obsolete in the coming years, which could mean that their devices could become worthless. No one wants their refrigerator to become unusable a year or two after purchase because the manufacturer is no longer open for business.
Standards for data aggregation, networking, and communication will all help to determine processes for management, transmission, and storage of sensor data. Aggregation is critical because it improves data availability (via the frequency of access, scale, and scope) that you can use when analyzing. One concern that will make it harder to arrive at agreed standards within this field is the issue of unstructured data. Information within relational databases, called structured data, can be accessed via SQL. However, the unstructured contents within NoSQL databases are not accessed through one standard technique. Another issue is that companies may not have the skillsets they will need on-staff to be able to leverage and maintain cutting-edge big data systems.
One of the key reasons that standardization will be so helpful to the IoT is simply that it will make everything easier – as noted by Daniel Newman in Forbes. Currently, you cannot simply plug in a device. Instead, apps and drivers have to be installed. The technology should be simpler. Through APIs and open source technologies, IoT manufacturers will be able to integrate their devices with the worldwide ecosystem that already exists. “If these items use the same ‘language,'” said Newman, “they will be able to talk in ways they—and we—understand.”
Derivation of Insights for Intelligent Action
Finally, the IoT must have takeaways. Cognitive technologies are used in this setting to improve analysis and spark more powerful findings. Key trends related to this field include:
- Lower cost of data storage: The volume of data that you have available will make it easier to get the results you want from an artificial intelligence (AI) system, especially since storage costs are lower than in the past.
- More open source and crowdsourced analytics options: Algorithms are developing rapidly as cloud-based crowdsourcing has become prevalent.
- Real-time analytics: You are able to get access to data that impacts your business “right now,” with real-time analysis through complex event processing (CEP) and other capabilities.
High-Performance Infrastructure for IoT
The above challenges are certainly not holding back the forward momentum of the internet of things. As it expands, strong and reliable cloud hosting will be fundamental to the success of individual projects.
Are you in need of a powerful cloud to back your IoT system? At Total Server Solutions, we engineered out cloud solution with speed in mind, and SSD lets us provide you with the high levels of performance that you demand. Get the only cloud with guaranteed IOPS.