A second problem is the connection itself. There are many ways to make the connection and that’s a problem, he says. “You have to choose what technology you want to put into the device, and if you don’t know which one, that can be a problem. After a few years if that system is withdrawn or stops working, or doesn’t have the required range or battery life, you are stuck.
The third problem is there are so many different areas to change the whole business process to allow the data to be used and processed. “You need different kinds of people, more digital skills, and change for partners with overall different business systems. A lot of things need to be done before you can introduce IoT and make it work.”
He points to a number of promising new developments but they need to treated with caution.
“AI is enormously valuable. In the IoT world the sensors will generate a huge amount of data ad all of this can move into a big cloud system and AI can act on that data and be really powerful in generating insights. That only works when you have got the data in the first place,” he said. “AI enhances an IoT system that is already there and working well – until you have that in place you can’t use the AI. We can rely on AI to deliver even more value from IoT once we hit those 50bn units
5G does not add any new IoT technology for sensors, he says. Those will still connect via 4G narrowband connections, and the option 5G may confuse as people may expect new capabilities. The low latency provided by 5G is a small subset, maybe 1 percent by volume, although by value it may be higher than that, he adds.
“I would concentrate on segments and sectors where IoT can be introduced where things don’t need to change too much with an immediate financial benefit. It needs to be packaged through systems integrators that can make everything work. It also needs to be done patiently,” he said.
“It’s very easy for the technology department to prove that the technology works, but a trial needs to be sponsored by senior management and the finance department. Otherwise you end up in the proof of concept hell – its easy to have a trial but very hard to move beyond that trial,” he said.
He points to Smart Cities where there are hundreds of trials, and a lot of data is being generated, from sensors in transportation, healthcare, metering, water and monitoring pollution levels. “All this feeds back into a data lake. Cities have changed dramatically with Covid-19 – there’s far less congestion, but the emphasis on tracking people to track their contacts and make sure that logistics work really well when people start panic buying or the need for protective equipment expands,” he said.
“This will only succeed if we bring everyone together, including local councils with concerns such as privacy. It’s about making sure that we find a solution across all of the partners, and that’s really difficult.”
Last year Webb published a book on the challenges of the IoT
Related IoT articles
- $50m fund for startups in Europe
- Bosch rolls out AIoT for its factories
- 5G edge AI computers use Tiger Lake processors
- NXP shows first details of edge AI i.MX9 processor
Other articles on eeNews Europe