We recently ran a webinar on Tech Trends 2021. We had some fantastic questions from our audience, some of which we didn’t have time to cover in the session so we got in touch with our experts and asked if they could answer them for our blog:
Would already existing IoT devices have to do any tweaking to make best use of 5G or would it work seamlessly in optimised way?
Chris Pont – Unfortunately, because the radio equipment differs, the devices will need updating to use the 5G network. They would benefit from the increased bandwidth, reliability and lower latency. The architecture of the system would need to acknowledge the processing that could be performed at the mobile base station to take advantage of the edge compute capabilities that 5G can bring.
How do you see 5G & enhanced connectivity benefiting manufacturing & engineering firms in East Anglia?
Chris Pont – Many of the big applications of tech in East Anglia are related to Agri-tech. Technology is helping agriculture by allowing farmers to make more informed decisions about where to place certain crops and exactly when to harvest to produce the best yields. By having more connected instruments and making use of 5G, farmers are able to collect this data and analyse quickly and efficiently using industry software. We’re also talking about big expensive equipment in many cases, which needs monitoring and maintaining. Self-driving tractors and harvesters are already being produced and 5G will help leverage that technology.
Lastly, engineering equipment often relies on firmware updates being rolled out. In much the same way as your mobile phone can update itself, these connected manufacturing, farming and engineering devices can update themselves over-the-air quickly and regularly. Obviously, there are safety considerations to take into account with how and when this happens.
Have you any insights on how AI is being applied to clients existing data, to cross reference and improve the quality, plug the gaps, fix data/ typing errors etc?
Andrew Burgess – Some of these technologies are not pure ‘AI” but do a good job of data cleansing by ‘fuzzy matching’ records (e.g. IBM = I.B.M.). Other technologies, some of which are built into data science and ELT platforms, look for missing data and can fill in the gaps with, say, average values. Also, some data can be inferred, such as age and gender based on other data in the record, such as name. Conversely, if there is too much missing data then the platform can recommend removing that field completely. And if some data is of the wrong format, e.g. some text data in a date field, then these can also be highlighted and, in most cases, auto-corrected.
Where AI really comes into its own is generating ‘synthetic’ data. Using a type of AI called Generative Adversarial Networks (GANs) the AI can generate data for another AI to be trained on.
One of the speakers gave an example of SARs in early 2000 and how technology was deployed to help. Why doesn’t the adoption of technology always mirror the speed of innovation?
Trevor Hunt – There are several answers to this broad question. Typically when new technology comes along there is an existing investment in technology that needs to be fully used before the latest tech can be deployed. In the past, tech spend has typically been a Capital Expense (Capex) and has to be amortised over several years for it to be affordable. Some call this Technical Debt. With the advent of Cloud we are now seeing services available on a pay as you go basis = Operating Expense (OpEx). This offers the ability to plug and play new technologies.
How do users cope with corporate hunger for benefits? New technologies and apps come and people get used to it and then big companies change privacy policies and users are in a compromising situation over whether to sign and continue using or compromise on data being shared…example WhatsApp.
Chris Pont – This is quite a big question on the subject of morals and ethics. If I’m interpreting this correctly, you’re asking about tying people in to a piece of software or a platform, making them reliant on it and then changing the terms? Often technology companies are doing this because they’ve attracted a wide user base and are looking to change their business model in a way to make it more profitable. Sometimes this is the only way that the platform will survive, and if users want to continue using the platform they need to make their own decisions around if they are willing to accept those terms. I think that the changes to those terms need pointing out to users clearly (not expecting the user to trawl through many pages and figure out what’s changed), Something like the financial services industry rolled out with “Key Facts” that do this well would help. There’s definitely some work to do here.
What one piece of advice would you give to entrepreneurs who are looking to venture into IoT and mobile apps development?
Chris Pont – There’s a low barrier to entry with some of these tools. The tooling is cheap or often free. A Raspberry Pi is around £15 for example. Microsoft make a lot of these tools free to allow you to develop on the .NET platform so my advice is to jump in and have a go! There are a lot of online resources such as Udemy taking you through the basics and how to get started.