Skills Shortage or Usability Crisis? Future-Proofing the Manufacturing Workforce - with Nick Leeder

Trend Detection Podcast ·

Show notes

Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – which gives you visibility and insights into all your assets, from single machines to full plants to help you reduce downtime, increase knowledge sharing and accelerate digital transformation across your organization.

You can find out more about how Senseye Predictive Maintenance can reduce unplanned downtime and contribute towards improved sustainability within your manufacturing plants, by visiting: www.siemens.com/senseye-predictive-maintenance

Transcript

It just shows you the mentality of how we think about selecting software. And actually they should be balanced because if it can do 4000 things out of the box, but nobody uses it, it adds no value, yeah. So, uh, you know, and I'm, I'm painting that as a very, very extreme picture, but the usability is fundamentally driving the adoption of these applications. Welcome to the trend detection podcast powered by Siemens, and now your host Nel Sullivan. Hello and welcome to the latest episode of the Trend detection podcast. Uh, really pleased to be joined by one of our regulars, one of our favorite regulars. I'm gonna add to that, er, Nick Leader, who you'll know from many podcasts, who joined our live event a few months ago, which was also published on here. Um, but you're back again, Nick, so welcome back. How, how have things been for having me. Good, uh, hopefully everybody had a good summer as well. I mean, it's, uh, now's the time of rest and recuperation really and back into the full swing of things, um, as we, as we hit the final quarters or uh the final few months of the year. Excellent, yeah, same here, same here, I have to say, but today we're talking about, um, a very important topic, and a topic that sort of emerged over the past year, 18 months, I guess, around sort of skills shortage, and, uh, you know, anyone industrial manufacturing space will know all about this, um, but we're, we're sort of framing this episode a little bit around, um, a blog you did for the World Economic Forum. Um, which we can of course link to in the show notes as well afterwards, but maybe you could sort of just open the conversation, Nick, just explain sort of the background to the post and the background maybe to this episode as well, like what, what are we're gonna sort of cover today. So, um, I, I think I guess the background was I hear the word skills shortage a lot. And um and it's undeniable that the workforce is aging in the manufacturing sector and more and more people are retiring or coming up to retirement and there is a real challenge of attracting um the younger generations into the industry. Um And a lot of it is focused around, oh look, we don't have, people aren't interested in manufacturing, which I kind of I kind of understand because there's a legacy of uh kind of being a bit of a dirty or conceptually being a dirty industry which isn't, it's a false idea. But actually what I was also looking at and grappling at is are we making it easy to attract the talent. And particularly when it comes to the applications that we use within the manufacturing working environment, you know, ERP systems, MES, predictive maintenance technologies, those sort of things, and looking at that and going, well, are these actually helping attract talent or actually are they a barrier for the uh bringing these skills or new skills into it. So do we, do we have a skills shortage? Or do we have a usability crisis? um, and I, you know, I frame that question in black and white because I think, um, having used many different applications over the years, I don't think it is quite black and white and I'm sure we'll go into that as we go through today's conversation. Yeah, and it can, it attracting talent because you'd, you'd, you'd imagine with all, you know, this new technology, you know, AI, we know about AI IoT automation. It's, it's quite good in, on the face of it, maybe a little bit of an outsider perspective for me, it's a really cool space to be in, but is that not reflected in, in reality, let's say. Well, cool, cool and manufacturing kind of go together when we talk about manufacturing technology as in the art of producing things. But cool and manufacturing doesn't really go always because I, I wrote a blog, a blog a while ago saying there's no fashion in manufacturing. You know, there's a reason why we still use PLCs from the 1980s on the shop floor because they work and they haven't broken. And until they break, nobody's actually gonna do about it, you know, manufacturing is about producing things profitably, safely, in the right quantity at the right time, and anything that affects risk around those things, um, are, are, are kind of is generally not looked upon positively, so. Um, when I look at these kind of cool, and I have to do the air quotes, and I apologize for that, but cool technologies, um, manufacturing is not always the fastest to adopt them. Um, and when they do, the applications I see have typically been built around an engineer thinking about the engineer's problem. So they're very complicated. Uh, and we see this in all sorts of applications. I said, ERP is a great example of it. Sometimes MES, if you're an engineer in computer aided design and product life cycle management solutions as well, they're built for engineers. And that means that they, they create very long learning curves to use them, um. And you know, many of the buyers actually post purchasing these softwares kind of regret the software decisions because they've, they've looked at the functionality tick box but not the usability tick boxes hard enough. And that means that implicate the implementations slow down, become difficult, and there becomes a usability debt that they create within the organization cos people are actually unwilling to use the tools because it, we make it difficult. For them to use the tools. So cool technology is great as long as the way we interface with that cool technology is equally as great. But manufacturing is slow also to adopt the cool technology. It's, I've, I've heard that term before, the, the software regret. It might have even been in a conversation with, was it Gartner or someone like that, this term of software regret, so that's, that's a real thing, right? You know, where they purchase it for whatever reason, and then the reality of the situa, or they, they've overlooked certain areas, which, um, which means in reality, it's not quite what they thought it would be. Yeah, well, it, it, you know, it's a great example, um, you know, I, I, I've worked on an RFP recently for an ERP system, I, I know, a, a full enterprise ERP system, and we are talking probably 4000 functional requirements. Can it do this? Can it do this, and less than 50 requirements on how it does it, how it interacts with people. Yes, that's, that's really interesting, Nick. So you're also, yeah, so you're saying usability is sort of at the core of the, these adoption challenges at the heart, you know, whether it's AI, whether it's IOT, whether it's automation, it's, you, you see it as overlooking the, the usability element of this is core to this issue. Yeah, and and and that usability piece um kind of is also emerging over time and different ways of interacting with the software. Um, So I think first of all is making sure that the software is intuitive, and now it's a very vague broad term intuitive, and it means one thing to one person and another thing to another person. But actually we have to step back and say what is the person trying to do with the job? Are they presented with all the information that they can do the job effectively and in a way that they can interpret without complex training. Uh, and knowledge And I think that's, that's really, really important and that's where I see user interfaces and user experience. Now I hear UI UX kind of statements fall down massively and there's some really interesting examples where companies, by the way, do it amazingly right, but at the same time, amaz, I've seen other examples um where they've done it amazingly wrong as well, and I think taking that perspective of the lens of the user rather than the engineer is really, really important. And then the second thing I, I think is. Generationally we are interacting with computers and devices in different ways. Um, you know, if I look at me, I still have a a laptop, a screen, a keyboard, a mouse. I actually have a telephone on my desk with buttons. Um, not, not the circular dials that we used to have that rotated around, but it's a voice over IP, but it's got buttons. Modern generation, you know, kind of live in a social media environment with interactions through mobile devices, for example. So I'm, I'm not saying this is the panacea to the problem by the way, but what I am saying is generationally the way we interact and process information is changing, yet we still seem to have user interfaces which are essentially an evolution from green screens. And green screens, by the way, if you're not of the generation, are how old mainframes used to, you've probably heard the term mainframe, which is a pre, you know, goes after punch cards, um but mainframes and the screens we used to have back in the 70s and 80s, uh that even that for me is early, by the way, um, were kind of very green fluorescent, so you only have one color, nom, no, and very poor resolution, but essentially a lot of what we have today is. You know, an evolution of that rather than completely rethinking about how modern generations are interacting, and that comes back to the whole skills shortage versus usability crisis. Yeah, and this, I guess the, the question that's in my head, it might, it might seem a bit of a strange question because we're talking about how to, you know, the, the skills, you know, skills shortage and usability and people actually physically using various apps. But, um, one thing that's sort of come up in various conversations I've had is around this, this idea that the, the users on the shop floor or whatever. Aren't necessarily the people using the app, but maybe they're looking more for vendors like Siemens or others to sort of manage the service for them. Is that, is that something you're seeing more and more that's, um, coming in conversations you're having? I, I'm gonna try, I'll need to unpick your question a little bit and ask you some questions on the basis of that because what, what I see in that is from an IT or an IT perspective at times, but mainly IT perspective, there is a, An increasing desire to outsource, particularly as we evolve to SAS or software as a service. So outsourcing all of that support function. But um. There are a there is a bit of a trend in the industry of um increasing engineering capabilities beyond that of just IT. Um, in fact I've, I've been watching with great interest in Accenture's Industry X practice as that evolves and how the businesses they've been acquiring. But actually, I, I, I think there is to a degree that, and I'm assuming that's where your question was going, by the way. Yeah, yeah, yeah, exactly, exactly. Um, I see that to a degree, but actually if you think about it. The core of what a business does is design, make and sell things in manufacturing, yeah, not all industries work on that same premise, but certainly in manufacturing you, you have to design and that involves creating unique IP. And solving a problem nobody else solves. Then you have a manufacturing piece which you can do that you can make that solution in the most profitable, efficient way and fulfill the demand of your customers effectively. And then you have the ability to sell that and actually attract those customers in the first place. Now there are elements within those processes which are always going to be non-value add. And I think those are the ones that are ripe for AI or outsourcing to be done. Um, you know, looking at having actually change analysts in the business, looking about the amount of engineering change that they're going through, that's not hugely valuable. Executing those engineering changes, finding solutions to those problems is, and I think that is the, the to me mentally where the cutoff between outsourcing and in and keeping things inside of the business sits. Yeah, well, it's like, I guess it's like, like anything really, you have to consider, you know, balance what's, I mean, I could talk about myself in a marketing context, you know, what to outsource and what to keep, um, you know, directly with, you know, my team as an example as well. So it's a, it's sort of a, a comparison to make there as well. Yeah, absolutely. Um, but, uh, switching back, so I, that was a little bit of a tangent off because we are talking about, you, you know, we. Yeah, but it was just, it was just something on my mind as I was thinking about this is, because, of course, we still, you know, you know, ideally, you know, in an ideal world, you would want people using apps on the shop floor and upskilling themselves, you know, both for their benefit, vendor's benefit, everyone's benefit, right? I think it benefits a lot of people. But I guess there's a responsibility with vendors of, of software, whoever, whoever it might be to. Evolve their designs to think about the human beings, the users at the end of it, so maybe you could you talk a little bit about that and how they can make these, um, you know, their softwares and platforms more, more intuitive to these users. Um, I, I think first of all, it's putting themselves in, uh, the shoes of the user, um. I, you know, and I, I'm gonna be slightly pokey and assumptional in some of these statements, but I, I can be shot down and I'm happy always to be shot down when I'm incorrect, but it'd be interesting to look at the major software vendors in the industrial space and actually look at their product development teams and the guys that actually build or or the people that produce the UIs and the UX's. How much time have they actually spent with end users on the shop floor, understanding what they're doing and how they're using it. I'm sure they've got the data in terms of how they use the application from afar, but there is a gap between. Kind of movement of mouse, if you want to call it that, whilst using the application and actually fundamentally understanding what an operator is doing and the decisions that they need to make and the time frames they need to make those decisions in. And I think that really fundamentally changes the game because if you can switch from theoretical design to actually almost going back to square one and saying what is it you're trying to do? I mean I, I, a really good example is uh often with projects we replicate the reports that exist today. What we don't do is actually go back and say what decisions are you looking to make from these reports and when. Then you'll actually find if you switch that question, I guess 50% of the reports that are produced aren't actually used. And the remaining 50%, only kind of 10% are actually fit for purpose. The other 40% are probably people having to look at that and then reinterpret the data from those reports. And user interfaces are no different, so I think, you know, we really need to look at this critically and say what are the decisions people are making and how do they make it. And with the advent of things like AI we can actually use this predictive position to provide much more contextualized and intuitive information to the user rather than having that linked behind tables and complex screens. Yeah, yeah, and that's, I don't, we've mentioned AI already a few times, but G AI as well, because I completely take that point on board. And I think, you know, as the example with Sensai, how, and I think of it from my point of view is like, I'm not, not someone who works in industrial manufacturing environment, but when I see the evolution. of the app from, you know, you log in, cases are raised, you can see graphs and, and plot, you know, what's happening, what trends are happening and, and, and why. But now we've got Gen AI where it's actually providing that contextual information. Do you see, it's a long way of asking, do you see, you know, this sort of in written form and easy to understand natural language is bridge going to help bridge this gap, you know, that we're talking about today. Yeah, potentially, because, um, I, I, I'm a, I'm a visual learner, so you know, you, you give, you give me four kind of extremes, one being a written report, one being, uh, a, a chart, one being simple red, amber, green, and something where I'm able to ask a question. There's a sliding scale of capability there and the way that I can interpret information. Um, But interestingly Gen AI unless it produces the chart, cos visually you can see and understand chart data very, very well, which you could probably never see in a table. So whether Gen AI gives you the answer is, you know, how are things trending over time. Now, the, the best one is red, amber and green and then that allows you to go deeper into the analysis. Um So the Gen AI piece is very, could be potentially very, very powerful, but all those statements rely on a trust in data. And What we still I think have fundamentally. Is that reports, let's face it, Excel, yeah, we tend to trust that more than what AI is telling us. So I think fundamentally is, and you know this is a interesting conversations I do have with manufacturers currently is they're looking at at AI they're not looking at creating a solid data uh and contextualized data um component before jumping into the AI question. And actually what you're gonna have is a um the classical sewage pipe problem is depending on what you put in, you get out. So, um. I think there's huge potential. But in manufacturing I believe still we have that data question to solve and contextualize data because sometimes the data without that context, you know, looking at the energy usage of a motor without actually knowing what you were producing at that time can be very false data, yeah, vibration, yeah, using that, so that the data points and relevancy coming together and that's where AI could be incredibly powerful, but we need to get that contextualized. Yeah, it, it's some, it's something we often say, I mean, it's interesting you mention Excel because it is amazing how, you know, probably still amazing because you can't make presumptions about, you know, what tools or whatever people use, but there is still quite a large majority still use Excel, right? And is that due to a lack of trust in, not just AI, if we're just talking about pre-AI software, you know, systems themselves, and this is a tried and trusted way. Is that sort of the mindset that's maybe holding Back and embracing some of these technologies. Um, so number one is where I see Excel being used is because there's a gap in the information they need to make decisions. So we go all the way back to what I talked about. Do we understand the decisions people need to make, and that's genuinely why they use people, a lot of people use Excel. It's, it's actually bizarrely, whilst being quite a complex user interface, something we've all grown familiar with. And we all know that in every business there are people who can just basically do mathematics in in Excel. There are slightly what we let's call them advanced users that can use V lookups and things like that. But then you get these other people who are amazing. They pivot table, they do pivot table calculations. They're, and they are able to answer questions that nothing else is answering. And they can uniquely play for it. This is the, I, I guess where Excel probably, no pun intended, where Excel excels is, um, I, I used to work with, um, uh, somebody, a VP of markets, um, in a company, and they used to talk about the sniff test. Does it pass the sniff test? And with things like, um, Excel, you can go and drill, interrogate the data so that you get a confidence, yeah. That it's passing the sniff test. So I think also that you kind of. That's I think is why it's used, it, it can answer questions that aren't being answered, which goes all the way back to the usability question at the beginning. But also it allows you to interrogate the data to make sure that your confidence gets high enough, and again that, let's call that the sniff test for want of a better description. …

Full transcript (WebVTT)