Author: Jen Berry for MobiDev
The Industrial Internet of Things (IIoT) goes beyond proof of concept and into wider adoption and deployment across a whole host of industries. A survey by IIoT giant Microsoft found that 85% of companies have at least one IIoT use case project. This number will increase, as 94% of respondents said they will implement IIoT strategies by 2021.
In fact, IDC predicts that spending on IoT technology will reach a trillion US dollars by 2022. The largest sectors investing in the emerging technology include discrete manufacturing, process manufacturing, transportation and utilities.
The growing applications and use cases of IIoT offer tremendous opportunities for global industries. As the technology matures this year, here are the three main trends that will reshape the industrial IoT landscape.
Efficiency IIoT trend: Wireless connectivity
IIoT refers to interconnected industrial assets and sensors over a wireless network. The lack of a catch-all networking solution remains the biggest challenge for its wide adoption.
There are several recent innovations that are being implemented to drive the growth of viable IIoT cases and improve operational efficiency.
The adoption of 5G will revolutionize connectivity for IoT-enabled industries. This nascent connection technology is tailored for IoT’s connectivity needs and is expected to catalyze its productivity.
Long-range low-power wide-area network (LPWAN) technologies like NB-IoT, LTE-M, LoRa and Sigfox are also driving innovation for IIoT connectivity. LPWANs for IoT sensors allow low powered devices to stream packets of data wirelessly but with a wider area.
What does it mean in figures? A new low-power LoRaWAN sensor may work during one year without battery replacement, compared to the current one to two months, and transmit data with up to 3 km coverage.
French-licensed LPWAN giant Sigfox services use cases for logistics and fleet management industries to improve traceability. Sigfox offers a much cheaper and more efficient alternative to cellular networks. Despite carrying only 12-byte payloads, Sigfox can sufficiently transmit GPS data in regular intervals — enough for fleet tracking and health info management.
IEEE is also developing a new Wi-Fi standard, called 802.11ax, which uses both 2.4Ghz and 5.0Ghz wireless frequencies. This gives IoT devices better access points with boosted capacity and bandwidth speeds as well as improved energy efficiency over previous Wi-Fi standards.
Among the rising amount of new protocols and solutions, how to decide which one of the innovations you should go with?
IIoT is not a technology solution. It is about solving a business problem, not adopting an innovation because of FOMO (fear of missing out). IIoT should not solve legacy technology, bloated database, security and so on. IoT can help you improve the quality of your product or service, reduce operational costs, optimize costly processes, or track assets. Every time we hold initial analysis for launching IIoT solution, we draw a possible technical roadmap. Still, those huge projects are often related to investments, risks and uncertainty. The optimal way is to start with PoC to test the idea and core field implementations.
IoT Lead Solution Architect
Business IIoT trend: Predictive Analytics
Data remains the backbone of IoT software and applications. IoT devices and end-points are growing in size and form. This enables businesses to find new ways of enhancing the interoperability of processes, people, assets, and data by leveraging new trends in AI and machine learning.
IIoT Data Analytics
While data analytics isn’t anything particularly new, MIT Sloan found that IIoT analytics will be vastly different from current data management paradigms. As counterintuitive as it may seem, data sharing among competitors will be the best way to make sense of this immense amount of information. To give you a peek into the future, Maryville University forecasts that by 2025 over 180 trillion gigabytes of data will be created worldwide every year. A large portion of this will be generated by IIoT-enabled industries.
The IIoT creates myriads of sensor data that enables an accelerated deep learning of existing manufacturing operations. The efficiency of an IIoT system increases if data scientists can analyze the incoming data and then operationalize those insights in a timely manner. For most AI based solutions, it’s not models or software that created a competitive advantage. But data. So, if there’s a way to aggregate unique data — it’s already half way to get business useful insights and operational optimisation.
Data Science Solution Architect at MobiDev
Similarly, remote data collection IoT platforms can wirelessly patch and update machinery in a timely manner through predictive maintenance. The intermittent access and quick response times this enables significantly decreases down time for machinery.
For example, deposits in heat exchangers are hard to detect and can clog the conduits. IoT devices that measure the temperature difference upstream and downstream, however, can issue warnings on anomalies that indicate potential blockages. Deploying Machine Learning-based predictive maintenance capabilities can reduce downtime by 20–50% and costs by 5–10%.
Tech IIoT trend: Edge computing
IIoT companies are now shifting implementation models towards edge computing. This technology allows data to be processed near the IoT devices, which reduces latency and the use of bandwidth. Edge computing enables the viability of everything-as-a-service business models and microservices-both of which rely on lightning speed computing capabilities and responsiveness.
The OpenFog Consortium which includes CISCO is committed to push fog computing to overcome the current challenges in edge computing-device management, scalability, and cybersecurity. As opposed to edge, fog computing can still process data in conditions where bandwidth is unavailable.
This means manufacturing devices connected through the fog can process data locally while transmitting only pertinent data with very small power consumption. Collision warning technology is also very compatible to fog computing. Vehicles with fog computing devices can also communicate with each other more efficiently and faster.
However, IEEE senior member and professor of cybersecurity at Ulster University Kevin Curran notes that fog computing introduces more cybersecurity risks to devices and virtual machines. Addressing these threats will be crucial for wider scale adoption.
Growing use cases
Industrial AI. Artificial intelligence programs built on the edge or on VMs enable predictive analytics. With the remote access capabilities of IoT and near-sensor machine learning combined, response rates and flexibility will reach unprecedented levels. Another emerging AI trend is leveraging deep learning and computer vision in AI visual inspection systems for detection and quality control.
Nakayama Iron Works is now reaping the benefits of their IoT-enabled and remotely accessed rock crushers and machineries. By leveraging AI-powered IoT and the eWON Cosy connectivity, the company can anticipate and quickly troubleshoot problems remotely. It has greatly reduced their operational costs.
The more the fourth industrial revolution continues to disrupt industries, the more IoT development companies will bring IIoT use cases to life. The maturing technology is poised to solve some of the most pertinent business problems today.
Full article originally published at https://mobidev.biz.