In theory, AIoT (Artificial Intelligence of Things) should combine the best of both worlds: the millions of connected IoT devices that gather real-time data, and the capability of AI to sort through this data and glean meaningful insights. In practice, this can be harder, and making sure that data gathered is diverse, secure and regulated is important to remember, noted Fatima Elleouet, Head of Global Vertical Markets at Alcatel-Lucent Enterprise in an exclusive conversation with IoT Insider.
The advantages and drawbacks of AIoT
AIoT refers to the convergence of AI and IoT technologies. Applying AI to IoT, Elleouet explained, “allows data to be computed in real time, enabling data-driven decision-making to accelerate interactions with the ecosystem, boost operational efficiency and improve security.”
Zeroing in on this more, the adoption of IoT devices across industry verticals – including industrial, healthcare and consumer – has led to “an unrelenting growth of data”. Imagine the kind of data industrial sensors gather, or a smart wearable worn around your wrist.
Unrelenting is an interesting word to use in this context, I thought, but it signifies a wider issue that without sorting through all of the data generated, organisations can quickly become overwhelmed by it, thereby defeating the purpose of collecting the data in the first place – which is to gain meaningful insights, such as the peak usage hours in a smart grid, or the patterns of behaviour for an elderly person living on their own.
“The data generated through IoT needs to be collated from across diverse departments which typically operate in silos and use different platforms,” Elleouet advised. “For example, to improve teaching and educational programmes, data often needs to be gathered and evaluated from multiple subject departments, teachers and support staff, and other third parties such as social services.
“We are seeing increased demand from across verticals for disruptive solutions that can help to break down these data silos.”
The benefits of AI are acknowledged by Elleouet alongside the challenges, which she referred to as financial, technical, operational or ethical.
“Data privacy is an immediate concern, especially for organisations [working] in healthcare, government defence and education, and which deal with sensitive data,” she said. “The cost of investing in AI is [also] a challenge for many organisations.”
This demonstrates the variety and versatility of applications AIoT is useful for, as well as the scope of challenges facing the businesses who want to adopt IoT devices across different verticals.
In one example, Elleouet said: “Ports run and rely on technology – digital twin, IoT, drones – to improve their operations and become more secure and smart. However, challenges around data usage remain; for example, much more work needs to be done to ensure that the control over sensitive data as well as communications between companies are completely secure.”
The application of AIoT
Some examples of how AIoT is being applied to different industries across the board is through data management and optimisation; real-time data processing; but also network optimisation and predictive maintenance.
Speaking about real-time data processing, Elleouet said: “AI can process IoT data near the source of the Edge, which can reduce bandwidth usage. This is critical for applications requiring real-time decision [making], such as autonomous vehicles, or manufacturing.”
Elleouet explained that she saw manufacturing as an industry vertical commonly cited as one that benefited immensely from the application of AIoT, but that it wasn’t alone in the advantages it reaped.
“All industries are seeing tangible advantages,” she said. “The use of AI allows organisations to get a better understanding of customer preferences to provide a personalised experience. In hospitality, AIoT enables customised guest experiences, hotels can gain insights into guest preferences and deliver services that surpass expectations.”
Looking ahead
Elleouet said future applications for AIoT she thought would experience significant growth included for security and safety purposes, and sustainability.
“AIoT can identify and collect information from a call with a disaster victim, or from a video call to geolocate the location of the accident and provide support to the victim and others,” she said. “AI combined with digital twin technology will enable the construction of intelligent, safe and passive buildings.”
She acknowledged that AI was in its “infancy”, but that shouldn’t mean to assume every possible avenue has been explored. “We will undoubtedly seek much more from the technology,” Elleouet concluded.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.