Return to overview

What industrial trends are gaining traction in 2022?

The last year has been challenging for IIoT, with many companies facing supply chain challenges and lockdowns due to COVID-19. But it’s also been a time of innovation, with deep R&D in materials innovation and growth in virtual and machine learning applications.

Let’s take a look….

Semiconductor chip shortage

It would be hard not to think of trends in IoT without mentioning the semiconductor chip shortage. Its causes are complex but include a fire at Japan’s Renesas Semiconductor Manufacturing (who produce about one-third of microcontroller chips embedded in cars globally) and COVID-19 shutdowns also reducing manufacturing. Stay at home orders saw an increase in purchases of consumer goods like PCs and gaming devices. While most of the news has been around auto manufacturing, semiconductors usage is across all sectors, including manufacturing, smart homes, industries IoT, so the pain has been widespread.

A poll last year of 262 embedded device and connected product developers in March found that 60% of industrial machinery and electrical equipment manufacturers are now heavily focused on securing IC supply chains. 55% of server and computer makers said they struggled to maintain chip supplies.

The shortage has resulted in delays in shipping, and in the case of cars, vehicles shipped missing components. But there has been a positive. Materials innovation in semiconductors is thriving. The use of gallium nitrides (GaN) to replace silicon and germanium has gained traction and the use of photonic materials. Photonics has been chiefly the interest of academics but is gaining commercial interest. Compared to semiconductor chips, chips made from photonics are smaller in chip size, run three to ten times faster, use less power, and contain no rare earth minerals. Photonic materials are now being put into commercially available devices directly.

The intersection of IIoT, edge, and AI continues

The use of AI to process and analyse IoT data is nothing new. Still, changes to the technical architecture of devices are creating new opportunities, with increased interest in bringing data analysis and AI technology to the edge – the extreme edge. Machine learning is traditionally a high latency process where data is processed in the cloud. By comparison, Tiny Machine Learning (or TinyML) is a meeting between machine learning and embedded systems, making it possible to develop machine learning models that can be executed on small, memory-constrained, low-latency, low-powered devices like microcontrollers.

This creates real-time and power-efficient edge analytics at speed without needing to store data in a server, increasing data security. An interesting example is an app in Africa called Nuru, which uses Google’s open-source TensorFlow technology. Nura can access over 5,000 images of cassava plants affected by disease to guide farmers in predictive health. TensorFlow makes it possible to use AI on a mobile device, without any Internet connection, which can be vital in areas with poor connectivity.

A growth of interest in the metaverse and digital twin

One trend that we’ve seen an increase of in 2021 is the metaverse. Nvidia created Omniverse, a virtual environment for engineers in 2020. This mixed reality platform allows people in different physical locations to join collaborative and shared holographic experiences on many kinds of devices.

IIoT is no stranger to virtual applications such as digital twins, but metaverses are fundamentally different. Digital twins are designed to mirror an existing object or process by utilising sensors attached to be a real-world asset for data. For example, a digital twin can be updated via senses and connectivity to reflect the health and performance of a physical item such as a piece of machinery. By comparison, metaverses can be more fantastical as they don’t need to be tied to a physical asset or its data, although they may be something we see developed in the future. The other key difference is immersion. Digital twins can be utilised by simply looking at a desktop or mobile screen. In contrast, a metaverse is developed for use with a virtual reality (VR) or augmented reality (AR) headset.

The Seoul Metropolitan Government (SMG) is now entering the metaverse, creating Metaverse Seoul’, a virtual communication ecosystem for all areas of its municipal administration. A virtual office will enable citizens to meet with public officials in avatar to access city services and resolve civil complaints without leaving their laptops. I’m not sure what this means in terms of the hardware required by the user or why this is better than using Zoom, but it shows that the technology is evolving beyond gaming and entertainment. At any rate, it will be interesting to see what industrial use cases gain traction amongst the noise of metaverse over the coming year.