The Internet of Things of not about 𝐓𝐡𝐢𝐧𝐠𝐬, it’s about 𝐃𝐚𝐭𝐚!

Originally posted in GiT Switzerland Community Space.

The convergence of Artificial Intelligence (AI) and IoT is commontly referred to as 𝐀𝐈𝐨𝐓. Around 60% of IoT initiatives involve AI and machine learning (ML) (Lamarre and May, 2019). Let’s take a look at the main reasons for integrating AI in smart connected solutions.

#1 – Firstly, AI can help us make better decisions. Humans are limited by biological constrains that make it difficult to process large amount of information and think beyond local solutions. In contrast, AI can analyze vast amounts of data and consider distant, exploratory information to derive 𝐫𝐚𝐝𝐢𝐜𝐚𝐥 and 𝐝𝐢𝐬𝐫𝐮𝐩𝐭𝐢𝐯𝐞 answers to problems. This can support, extend, or even replace 𝘩𝘶𝘮𝘢𝘯 𝘥𝘦𝘤𝘪𝘴𝘪𝘰𝘯-𝘮𝘢𝘬𝘪𝘯𝘨. AI can use rule-based, ML-based or hybrid algorithms to support decision-making.

#2 – Secondly, AI can drive the development of novel products and services by adopting a 𝐜𝐮𝐬𝐭𝐨𝐦𝐞𝐫-𝐜𝐞𝐧𝐭𝐫𝐢𝐜 𝐚𝐩𝐩𝐫𝐨𝐚𝐜𝐡 that emphasizes service, which is often ignored in the traditional product-centered approach. To do this 𝐢𝐧𝐝𝐮𝐬𝐭𝐫𝐢𝐚𝐥 𝐩𝐚𝐫𝐚𝐝𝐢𝐠𝐦 𝐬𝐡𝐢𝐟𝐭, the human factor cannot be overlooked, as 𝘨𝘳𝘦𝘢𝘵𝘦𝘳 𝘤𝘶𝘴𝘵𝘰𝘮𝘦𝘳 𝘱𝘳𝘰𝘹𝘪𝘮𝘪𝘵𝘺 is one of the key value drivers of AIoT solutions.

#3 – The third reason is to optimize processes. AIoT can be used to address various use cases, including demand forecasting, product quality control, 𝘤𝘰𝘯𝘥𝘪𝘵𝘪𝘰𝘯 𝘮𝘰𝘯𝘪𝘵𝘰𝘳𝘪𝘯𝘨 and 𝘱𝘳𝘦𝘥𝘪𝘤𝘵𝘪𝘷𝘦 𝘮𝘢𝘪𝘯𝘵𝘦𝘯𝘢𝘯𝘤𝘦.

Let’s now consider some of the limitations and implementations challenges for AIoT.

In addition to the 𝘗𝘰𝘊 𝘵𝘳𝘢𝘱 discussed in my previous post, a key challenge is ensuring high-quality data, since ML algorithms rely on historical data for training. As the saying goes, “garbage in, garbage out”. Another pressing challenge is neglecting the 𝘩𝘶𝘮𝘢𝘯 𝘧𝘢𝘤𝘵𝘰𝘳, which can undermine the efficacy of even the most technically sound solution. This is why a 𝘗𝘰𝘝 is crucial, as it can help address the questions I raised in my previous post. 

Finally, in order for AIoT to be successfully adopted, 𝐭𝐫𝐮𝐬𝐭 in AI is crucial. This topic has gained increasing attention since the launch and popularity of ChatGPT. However, there is a trade-off between simplicity and accuracy in AI systems. Simple and straightforward systems may be easier for humans to understand and trust, but more complex neural network-based systems may provide more accurate predictions at the cost of less clear explanations (Gunning et al., 2019). To ensure that AI solutions generate trust and meet 𝐞𝐭𝐡𝐢𝐜𝐚𝐥 requirements, it is imperative to assess and guide each AI project with a 𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐥𝐞 𝐀𝐈 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤, like the one Dr. Lisa Falco developed together with her team at Zühlke.

Final thoughts 

When Operational Technology (OT) and Information Technology (IT) tie the knot, they offer valuable insights and substantial value. But how can we ensure that this union lasts?💍In this last post, we will explore some key success factors that can help create a long lasting marriage between 𝐎𝐓⚙️ and 𝐈𝐓 🌐, paving the way for a seamless integration into 𝐈𝐨𝐓.

IoT initiatives require a mastery of hardware, software and physics, employing a probe-sense-respond approach to 𝘵𝘢𝘮𝘦 𝘤𝘰𝘮𝘱𝘭𝘦𝘹𝘪𝘵𝘺. As we’ve seen, ideas are cheap – so be ready build, test and learn quickly. À propos 𝐚𝐠𝐢𝐥𝐞: be prepared to create and destroy. If your PoV fails to support your hypothesis, pivot or dump it, without becoming attached to it! Keep things 𝘴𝘪𝘮𝘱𝘭𝘦, particularly at the beginning, innovating from local to global, from simple to complex. Above all, create a 𝘤𝘶𝘭𝘵𝘶𝘳𝘦 𝘰𝘧 𝘭𝘦𝘢𝘳𝘯𝘪𝘯𝘨. Remember that adaptation takes time, so be ready for a learning curve. This final point Is not only relevant to IoT but essential for cultivating the 𝐜𝐨𝐮𝐫𝐚𝐠𝐞 𝐭𝐨 𝐢𝐧𝐧𝐨𝐯𝐚𝐭𝐞 in any domain.

Maintaining a 𝘤𝘭𝘦𝘢𝘳 𝘷𝘪𝘴𝘪𝘰𝘯 is crucial at all times. Since there is no off-the-shelf solution, foster 𝘢𝘥𝘢𝘱𝘵𝘢𝘣𝘪𝘭𝘪𝘵𝘺 of existing technologies to create value. Collaborate with partners to acquire the missing skills and expertise, while prioritizing knowledge transfer. Keep in mind that the foundation of a successful enterprise adopting smart connected solutions, or any other technology for that matter, is a loyal and highly driven team. Ultimately, 𝐩𝐞𝐨𝐩𝐥𝐞 are the key to success. 

Returning to the original question of the first post: remember technology is not the ultimate goal, but a toolset to solve a problem of the 3 holistic perspectives: people, planet and business. Start by asking the right questions: what 𝘯𝘰𝘯-𝘵𝘦𝘤𝘩𝘯𝘪𝘤𝘢𝘭 𝘱𝘳𝘰𝘣𝘭𝘦𝘮 am I trying to solve? What 𝐯𝐚𝐥𝐮𝐞 will it deliver, and to whom?

I hope you enjoyed the posts as much as I did, and I'm looking forward to hearing your thoughts on them.

PS: for more insights on this topic, I highly recommend reading the book “𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐞𝐝 𝐁𝐮𝐬𝐢𝐧𝐞𝐬𝐬” by Oliver Gassman (HSG), Fabrizio Ferrandina (Zühlke) and many other authors from the industry.