Artificial Intelligence is Making Waves within the Internet of Things Ecosystem

The Internet of Things (IoT) and Artificial Intelligence (AI) are common buzzwords in technology that are often closely linked, but it’s questionable how many IoT devices actually make use of AI today. A new wave of concerns must now be considered, including how do our fears over security impact on the IoT? In addition to this, what do design engineers actually think about AI? The answers lay within new research published by Farnell.

The research confirms that there is a strong adoption of AI within IoT devices, with 49% of respondents already using AI in their IoT applications, which is unsurprising given proven benefits of AI for businesses. Interestingly, Machine Learning (ML) has been shown to be the most used technology (28%), and I must admit, I thought this might have been slightly higher due to adopters of IoT using integrated ML and AI within their equipment to enable predictive maintenance solutions.  

As evident above, we can clearly see within the survey results that AIoT is a major trend, demonstrating the beginning of the process to build a true IoT ecosystem. The adoption of AI within IoT design is coupled with a growing confidence to take the lead on IoT development and an increasing number of respondents seeing themselves as innovators. But the survey also shows us that some engineers (51%) are hesitant to adopt AI because of two key reasons: firstly, it is new technology they are not familiar with and secondly they require specialized expertise in how to implement AI in IoT applications. All we need to do is look at the skills gap across many plants and factories currently, and it’s easy to see why over half of engineers share this reluctance. It’s not necessarily a fear of the unknown, but more the lack of knowledge to implement it. 

Security also appeared as a big concern, one that’s been examined previously, and no doubt will continue to be investigated as we advance with IoT technologies. Designers of IoT implementation need to consider security to protect company’s data. However, interestingly, only 35% of participants in the survey cited security as their biggest concern. I say interestingly because this is a reduction from 40% of respondents who viewed security as their biggest concern in the 2018 study. On the other hand, security is still ranked significantly higher than connectivity and interoperability due to the type of data collected from “things” (machines) and humans, which can be sensitive and personal, and thus a big concern. Many of our customers who are initiating new IoT projects treat IoT security as their top priority, and focus on the implementation of hardware and software security to protect against any kind of potential threat. Ownership of collected data is another important aspect of security we must understand. Unsurprisingly, 70% of respondents prefer to own the data collected by an edge device as opposed to it being owned by an IoT solution provider.  

As an engineer, I prefer to design complete edge-to-cloud and security solutions myself. I was delighted to discover through the survey that I’m not alone here. Many engineers (46%) prefer this approach, particularly if in-house expertise is limited or time to market is critical. An increasing number of engineers are now more open to the idea of integrating product ready solutions, such as SmartEdge Agile and SmartEdge IIoT Gateway, to provide complete end-to-end IoT solutions. This is further supported by an increase of 12% of respondents who confirmed that they would consider purchasing third party devices in 2019 than 2018.

A key trend that I found interesting in comparing the 2018 and 2019 survey results is that the growing range of hardware available to support IoT development continues to present new opportunities. Thirty-three percent of the respondents claimed that technology innovation is being driven by start-ups, which is an increase of 7% compared to 2018. I believe this to be significant from a distribution perspective as those start-ups are benefiting from the wide range and broad availability of modular solutions and single board computers in the market. To add to that, the number of respondents adopting off-the-shelf hardware has also increased to 54% from 50% in comparison to 2018.

In 2020, we are seeing opportunities within the IoT and AI growing and they will continue to do so. This will be fuelled by access to an increasing number of hardware and software solutions, enabling developers to bring products to market faster than ever before, and without the need for specialized expertise. As a result, we see new entrants entering the IoT arena which in turn gives more developers the opportunity to innovate and offer solutions that overcome challenges and improve people’s lives.

Our daily lives are quickly seeing the integration of smart devices, with many of us turning to AI in the home with commands, such as “Alexa, play [insert favourite band] on Spotify” or “Hey Google, how long does it take to cook roast potatoes?”. Amazon’s Alexa and Google Home are common evidence of the widespread adoption of IoT in the consumer space. Yet, despite this, the 2019 survey saw a slight shift in focus away from home automation. The number of respondents who considered it to be the most impactful application in IoT within the next five years reduced from 27% to 22%. It would seem that as an industry, we believe the focus will move from the consumer space to industrial automation and smart cities. Out of the respondents of the latest survey, 22% and 16% respectively, view industrial automation and smart cities as rising interests. I for one happen to agree with this perspective as the value that IoT data can bring to operations is phenomenal. It is something we are seeing a lot of in industry where more facilities are converting to full or semi-automation in robotic manufacturing and increasing investment in predictive maintenance to reduce production down times.

The survey data will likely change as we move through 2020, embracing new technologies and increasing our general awareness of the role AI can play in manufacturing. The Farnell IoT survey was conducted between September and December 2019 and we saw 2,015 respondents participating from 67 countries in Europe, North America and APAC. Responses were predominantly from engineers working on IoT solutions (59%), as well as buyers of components related to IoT solutions, hobbyists and makers.

You can review the full survey results, here.

.