They are impressive individually. The Internet of Things (IoT) has been around for several years to anticipate traffic problems, gauge harsh weather conditions, deploy and control intelligent trash cans or help you with your supply chain. Now, with increasing data volumes, Artificial Intelligence (AI) is often seen as the panacea… to make sense of everything. So, it’s easy to see why the two technologies mesh so well together. But sometimes, it’s also hard to identify suitable use cases delivering tangible business value. How can companies elucidate what’s worthwhile?
It’s clear that they were meant to work together. The growing stream of data generated by connected devices —which is forecast to reach 79.4 zettabytes by 2025— needed a technology to efficiently sort everything out and help companies reap benefits from previous investments. AI has come to the rescue. Now, IoT Analytics assesses that the adoption of AI-powered IoT projects will jump from 10% to 47% by 2027.
The combination of both technologies is key to developing smart solutions for smart cities, the health sector, leisure and tourism, agriculture or banking and insurance, not to mention transport where cameras and other connected elements can make self-driving vehicles make autonomous decisions in real time.
“IoT allows us to capture a lot of data from many connected devices at distant points. Depending on the specific application, this data can feed AI systems that can be used, for example, to train models, make predictions, or make recommendations, among other things,” says Joan Mas, Scientific Director at the Digital Area of Eurecat, the main technological center of Catalonia.
In short, AI tells devices how to better distinguish and react to the information and requests they receive. Therefore, AIoT sensors throughout a manufacturing facility can help monitor machinery for signs of failure, or find inefficiencies in your operations, excessive energy consumption and even discover better logistics for quicker deliveries.
Starting with pilots
Yet adopting AI —and particularly GenAI, which transcends simple data processing— “is not the magic wand for every enterprise,” says an industrial AI expert who participated in IOTSWC24. “Enterprises need to strategically evaluate GenAI’s value by defining clear objectives, assessing their data and AI readiness, the ethical implications of AI in their industry and the cost of ownership to drive the ROI,” he adds.
So, what to do in this scenario? “Starting with pilots, to experience the outcomes firsthand can keep a good handle on the GenAI journey. Also, staying informed about peer-industry trends can help enterprises decide whether generative AI is the right investment for their domain-specific problems,” this expert recommends.
“There are many solutions that can be tailored to the needs of each business. But it’s clear that it will be necessary to invest in computing infrastructure, —most likely in the cloud, while taking into account cybersecurity criteria to protect data— and also in the AI tools available on the market that best fit your needs. And when market solutions do not respond, companies can always turn to engineering firms or technology centers to adopt unique and innovative designs specific to the problem they have to tackle,” says Joan Mas.
Given how young this technology is, being cautious is a good thing. However, companies that aspire to be pioneers in their industry segment will need to accept a certain level of risk to have a competitive edge over their peers. To ensure their path, they can collaborate closely with Gen AI technology vendors to speed their search for groundbreaking solutions relevant to their sector.
Even so, some people will still have doubts throughout the process, as is common in any relationship, partly because of their business model, partly because they may have the hypothesis that AI reasoning might be done somewhere else. And that’s ok.
Be that as it may, the IoT and AI blend will undoubtedly continue to advance. But at this point it’s important to bear in mind that “without the human in the loop, technology isn’t so great”, as Jessica Poliner, CEO of Relayr, put it at Internet of Things Solutions World Congress 2023.
Sean O’Reagain, Deputy Head of Unit Industry 5.0, Directorate General for Research and Innovation, at the European Commission, went even further: Now the industry needs to “achieve societal goals beyond growth and be more human-centric”, and that goes beyond being able to use AIoT or Gen AI. The true question is not if we will use it, but for what purpose…
Article by: Anna Solana