Beyond 1% - Industrial AI for a better future? - a panel discussion
Show notes
Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – the platform which enables predictive maintenance at scale across all of your assets, across all of your plants.
In this episode, we discuss:
Summary:
How AI is transforming industries beyond hype: Discover real-world examples of AI’s impact in industrial settings—from predictive maintenance at enterprise scale to visual AI for infrastructure monitoring.
The link between AI and sustainability: Learn how AI isn’t just a tool for efficiency, but a key enabler of sustainability, driving carbon and cost savings by extending asset life, reducing waste, and avoiding unnecessary inspections.
The “Beyond 1%” philosophy: Explore how Australia can play a far greater global role in decarbonisation by leveraging digitalisation and AI—not just focusing on its own emissions footprint.
AI’s double-edged environmental impact: Understand both the environmental costs of scaling AI (like energy use in data centers) and the innovations that are helping to mitigate them through responsible infrastructure and smarter operations.
Why data and vision matter: See how tapping into underutilized assets like camera networks and drones enables smarter, scalable monitoring—helping companies extract insights without adding hardware complexity.
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
Unplanned down time is cost avoidance. You never plan to spend it. But as you get more and more into understanding and doing this at scale, and when you start to look at an entire enterprise and all of the machines, the level of data, uh, data, that's, the level of data and, and the, the, the things that you can explore there, you start to be able to say, well, actually I don't need to perform these inspections as often, I don't need to go to the site as often. I can actually now, uh, turn my Capital acquisition of this machine. Into, into a longer term payback because maybe I only expected this machine to last 7 years, but actually with the information that I'm looking at through predictive maintenance, I have the confidence now to say, well, I can extend the lifetime to 10 years. And so you're turning the conversation of sustainability and predictive maintenance from a cost center into a real cost saving. That's money that you can put in your pocket. Welcome to the trend detection podcast powered by Siemens, and now your host, Nel Sullivan. Good afternoon. On behalf of Engineers Australia, I'm delighted to welcome you all to today's webinar, Beyond 1% Industrial AI for a Better Future. My name is Amanda from Engineers Australia. Firstly, in keeping with our custom, Engineers Australia acknowledges the traditional custodians of the country throughout Australia and recognizes their continuing connection to land, waters and community. We pay our respects to them and their cultures and to elders, past and present. Before we get started, I'd like to acknowledge that today's webinar is being hosted with Engineers Australia's industry partner, Siemens. Siemens is a global technology powerhouse that started its Australian operations in 1872. Celebrating more than 150 years of operations in Australia and more than 175 years of operations globally. Today, Siemens brings together the digital and physical worlds to benefit customers and society. Active around the world, the company focuses on industry, infrastructure, mobility, and healthcare. This includes intelligent infrastructure for buildings, distributed energy systems, and automation and digitalization in the process and manufacturing industries. Creating technologies for more resource efficient factories and resilient supply chains, smarter building and grids, cleaner, comfortable transportation, and advanced healthcare. The company empowers customers to modernize the industries that form the backbone of economies, transforming the everyday for billions of people around the world. Today we will hear from our expert speakers in an insightful panel discussion. And I'd now like to welcome our moderator for today's session, Keith Ritchie, Head of Communications and Government Affairs, Siemens. Keith is a thought leader, futurist, and storyteller. He has been instrumental in driving topics around the energy transition, green hydrogen, Smart-Infrastructure, digitalization, and industry 4.0, including his role in establishing the Prime Minister's Industry 4.0 task force. He believes that accelerating digitalization and technology is critically linked to achieving positive outcomes for business, the economy, society, and climate change. Keith is behind the visionary concept of beyond 1%, which explores how Australia could decarbonize the world well beyond our emissions. Please welcome Keith Ritchie. Thanks to Engineersers Australia for that great introduction. It's a pleasure to be here today, uh, on a topic which I think is, um, it's very sexy at the moment, this topic in fact. Uh, there's a saying in showbiz that sometimes, um, a great, a great artist gets a number one hit and says it took me 20 years, uh, you know, to become famous sort of thing overnight. And overnight success. So Siemens has been involved in artificial intelligence for about 50 years now, and the topic of AI seems to have really taken off, but for good reason. And today we've got experts with us from different industries and different backgrounds representing different opinions. So we're going to have a pretty free flowing conversation. And hopefully some engineers out there listening will, will get a lot out of this. I'll start with Rick. Rick Irons McLean is the the Chief Technology Officer of Microsoft. Um, I'll come back to each of you for a bit of an explanation and understanding of what you actually do. Alex Hill, um, is a founder of Sensei, which is an AI tool which uh Siemens has acquired, and Hanno Blankenstein is um also the founder and owner of Unleash Live, which is visual AI. But maybe let you guys describe what you actually do. Start with Rick. It's a, it's a really interesting thing. I don't quite know what I do half the time, to be honest, Keith, but, um, and also whenever you introduce people as experts, there's always setting a big expectation. But you start sweating. Yeah. So I'm the chief technology officer for Microsoft's, um. Enterprise commercial business across Australia, New Zealand, I'm an external facing Chief Technology Officer, so I spend all of my time working with customers and industry groups to really look at how technology like AI can be leveraged to deliver outcomes, because at the end of the day, AI is a tool, like other things like IOT and cloud, it's a tool. So for me it's how we actually get value out of that and that's what I spend most of my time doing. OK, Alex. So we started 10 years ago to develop AI to help monitor machines at industrial enterprise scale, so. Yeah, we can cover the 50 machines, the 100 machines, but that's traditional condition monitoring. In our frame of mind, when we talk about enterprise predictive maintenance, it's the whole enterprise, tens of thousands of machines, everything from lighting systems to a transformer to a conveyor to whatever there is. And AI is core to that because humans are not very good at monitoring tens of thousands of machines just by themselves. OK, hello. Thanks Keith, um, at Unleash we believe vision is actually your most powerful sense. So we are specialized to extract data from computer vision or through computer vision, which is one area of AI and we are connected into your traditional CCTV cameras, it can be drone cameras, robotics cameras. And we then use computer vision to really help monitoring of critical infrastructure, but also large scale infrastructure such as roads, power lines, wind farms, and in effect we supercharge your camera network to extract data for your maintenance and operational efficiency. OK. Rick, I attended, uh, last year in December, I attended the uh Microsoft AI summit in Sydney. And um there's a couple of things that really stuck in my mind and it, I guess it brought a a a light bulb moment on the sheer pace of change with AI. One of them was, I seem to recall your COO, global CEOO saying that um in 18 months, Microsoft has built a $10 billion AI business, basically. Is it, is that correct, something like that? Uh, I'm not going to go into the specifics of the numbers, but, but it's that level of pace and it's that level of output and, and revenue that we are, we are seeing. So, and I, I think that's something that's going to continue to grow exponentially, particularly if you think early on. We're still at the nascent stages of generative AI development, which is really what's accelerating the AI transformation, and we haven't yet seen this scaled out across the world, across multiple organizations yet, so I think we're going to continue to see that as AI itself, particularly Gen AI, becomes more and more capable. And it delivers more and more value and outcomes. So I think it's definitely a, a trajectory that's going up like that, Keith. Yep, yep. So, in, in the context of what we're here for today, this series that we've been running with Engineers Australia is around the topic of beyond 1%. And the 1% represents the greenhouse gas emissions that the nation is responsible for. Um, everything that we do, every commitment we've made, the net 0 2050 commitment, the energy transition, the greenhouse gas targets. All of that is to address the 1% of emissions that we're responsible for globally, and we believe actually if you accelerate technology and digitalization, including AI, Australia could contribute to decarbonising the world by a lot more than 1%. So that's what we're really wanting to explore today, and sometimes we have to unpack that to see and, and put a different lens over what these outcomes are, where our emissions sit. Um, what sort of things, you talked about predictive maintenance, Alex. How does predictive maintenance have any impact on sustainability? That's a great question, and usually in the market when people talk about predictive maintenance, they're, they're talking, talking about fundamentals of avoiding unplanned downtime. And sure, that's important, when you want to use your machines, you want them working, and it's a basic principle of if I can't produce my product and I can't sell it and I can't make money. But we find that a reasonably unsophisticated argument, unplanned downtime is cost avoidance. You never plan to spend it. But as you get more and more into understanding and doing this at scale, and when you start to look at an entire enterprise and all of the machines, the level of data, uh, data, sorry, the level of data and, and the, the, the things that you can explore there, you start to be able to say, well, actually I don't need to perform these inspections as often, I don't need to go to the site as often. I can actually now uh turn my Capital acquisition of this machine. Into, into a longer term payback because maybe I only expected this machine to last 7 years, but actually with the information that I'm looking at through predictive maintenance, I have the confidence now to say, well, I can extend the lifetime to 10 years. And so you're turning the conversation of sustainability and predictive maintenance from a cost center into a real cost saving. That's money that you can put in your pocket. It's money that you had budgeted for that you don't need to spend. And then you have all of the knock-on effects of, well, if your machines are behaving better, they're likely to use less energy, they're likely to break down less and spill something or cause some form of environmental damage, and all of these things knock up, but you need a very sophisticated system to be able to capture that information, process it and work with it. And humans aren't very good at that when it's at scale. So we, we started developing Sensi 10 years ago to, Really uh codify that and make it simple for humans, we talk about machines all the time, it's ultimately about humans, how can we make their jobs easier, how can we make this story better? OK. Hanno, give us, give us an example of where you talked about this vision being at the core and how does that work with AI? What, give me an example so I understand what you're actually talking about. Sure, and um being partners of Sensi, we see that it all starts with actually capturing really accurate and good data in the field at scale, and um you can do that manually, so at the moment they are, Many, many engineers and humans just collecting all sorts of data, infrequently, um, often on paper or sometimes they have an Excel sheet, um, it's siloed so it sits on their laptop. Um, so that is a problem because if you have 300 wind turbines or 750,000 power poles or large machinery spread across the world in iron ore mines, you struggle with actually understanding the performance, um, the, the lifetime of your assets, when maintenance is required, etc. So now clearly in the last years there was a trend towards IoT, so you put sensors onto machines or you extract digital data from your machinery, which is fine, but comes with a lot of challenges. You need to deploy hardware, you need to connect the hardware, you need to maintain the hardware so you have another asset in the field. And so with computer vision. What we believe is a very complementary data set to IOT and the manual based inspections is just putting Data into cameras, um, you see that, um, for example, all of the conveyor belts in the Pilbara, are actually connected with more than 80,000 cameras. Um, they are often used for security. And now we inject more insights into those camera systems. So now suddenly, Um, you can analyze things as depth, speed, heat, vibration. All of those things exist through computer vision and fascinate, it's fascinating that our eyes actually sense the same things that I just talked about. Your eyes can sense depth. Your eyes can sense vibration. And it is extremely powerful, so suddenly you don't have a single IOT sensor anymore to put a vibration sensor somewhere. You can just have a multi-sensor environment with existing camera systems. Now you can take that into further automation and say, why don't you just deploy it on moving cameras such as drones or robots moving through a factory or flying across transmission lines. And it becomes a sensor that is not fixed to a machinery in one place, but you actually can almost collect it. It's like a scout that collects the data. And so if you have that paradigm of scale. Automation and low cost data capture, then you can really power up tools like, um Azure Cloud or Excel or Sensii, which suddenly then produces much more insightful data, and it's not siloed anymore and it's not static anymore. I love the multiple use cases there of, you've, you, you know, people understand vision, they understand putting cameras places, there's a use for that already, you can make use of existing investments. And and take it to this much uh more interesting use case where you're talking about efficiency, sustainability, and all these other things with the with the technology that's really well established. It's already there, yeah. And when you look at the, um, I think the trend really accelerated um with the iPhone, right? Um, you can, you can think of the iPhone as a consumer device, but what it did was, The camera in phones suddenly made it accessible for anyone, not only engineers, but everyone understood the power of photography first of all for the consumer market, but then suddenly Apple added LDA as a complementary sensor. It added depth into the sensoring, so. It's amazing how when you couple software with the camera sensor, what is possible and previously with your traditional CCTV or your SLRs, you never had that power. So it's amazing how you can backend all of this with strong compute power in the cloud or on prem. And you get this rich multi-sensor environment. Yeah, before we kind of go off this, um, I'd like to probably address the elephant in the room first of all, because we're we're already jumping into a solution using AI and so on to solve sustainability. I think we have to look at this in two ways. We know that there's going to be an impact on sustainability, particularly environmental climate through the use of AI. And we, we can't get away from that, we don't know what that's gonna look like yet because we're not yet running multiple scaled AI workloads all around the world, so there will be an impact. There will be more energy consumption, there will be potentially more water consumption and so on, and then more materials as we build data centers, so we actually have to think first of all, how do we mitigate against the impact of AI on sustainability, particularly the environment. So like at Microsoft, when we build our data centers, they all run now 100% on low carbon energy sources. Um, any of our new data centers, the AI ones that come online, um, they don't use water for the cooling process. It's alternative liquid cooling, cooling plates, and so on. In the materials that we actually use to build it, we're actually using, you know, low carbon steel, uh, low carbon cement. We're even building our first data center in the US which is made of wood. So there's a whole load of things we can do. We also know that the components, server components and so on are getting more efficient over time, so they draw less energy. But there is going to be an impact, so we have to be responsible as organizations, particularly large organizations, to make sure we mitigate against that impact. However, we also know if you flip that around, that the only way we're going to solve some of the big sustainability challenges is through the use of AI and we can't get around that. So if you think about, you know, large complex modeling for the next generation power utility grid, how do you model what renewable energy looks like when it integrates? How do you actually operate that dynamically when you've got. Multiple different sources that may be going up and down at the same time. How do you predict climate change modeling, risk modeling? It has to be done via AI, to your point earlier, Alex, there's no human that can do this. So we have to think about leveraging AI for these types of capabilities and across use cases, AI can do this. So in the Climate Leaders Coalition of Australia at the moment, Microsoft are leading up a group on AI and we'll have two publications coming out in March and April. The first one will be how any organization can think about mitigating that impact. The second one is really looking at how we take AI, which is, as I said, a tool, and we start to map that to different industry use cases to say this is how we can solve some of those sticky problems in each industry which may be slightly different. So I think it's really important we take that responsibility first of all to say there could be an impact, we must mitigate. But then really flip this around and say, to your point, how do we not only solve Australia's problems, how do we go beyond that and look at how we have a better economy, how we can compete on a global scale beyond 1%. In some ways, what you've described is the philosophy of net zero itself. So philosophically, uh, to achieve net zero, you are meant to, uh, make your operations as efficient as they possibly can before you rely on offsetting. And that's, that's exactly what you've described. But I think that flip side is really important. Data centers is one thing, of course, you need more and more data centers, uh, to handle that huge amount of data, but we know that there might be developments there on the way the chips are designed and quantum computing and other, other things, uh, which will reduce the, the size required. The heat generated and so on, and the software engineering and the improvement in the models which we're already seeing, so you're absolutely spot on. So, so this is, this is an improvement that's, that will definitely take place exponential improvement because now you can use AI to improve the operations of your AI data center and achieve a level of granularity in that control and sophistication and efficiency that would drive a human team wild. Yeah. And, and I think this is a really important point about how, There are two elements here, right, and I might just be summarizing it in a different way, but there's AI and the capability of AI. We produce more we consume more energy through the use of AI, which is a concern, but then you have the energy production and the energy transmission and distribution. So it's it's almost like we can put lots of Teslas and e-mobility into our transport system. But unless we have the energy generation, not coal-fired anymore, um, there's no effect, right? So to your point, net zero really has to address the entire value chain of scope 1, scope 2, and scope 3. And um one anecdote or example, uh, I recently, um, had a conversation with the CEO of one of our clients, um, the transmission and distribution company. And through our drone flights where we completely automate the inspections, we save about 70,000. …