Podcast | Ep 80 | Innovation in operations: balancing risk, quality and efficiency

Maria Rampa: Hi, I’m Maria Rampa, and welcome to this episode of Engineering Reimagined.

Have you ever seen or used a new technology and wondered: what problem is this actually solving?

That’s the first question Aurecon’s Chief Operating Officer Scott Powell asks when he’s presented with new technology, and it’s at the heart of his conversation with Professor Marcus Holgersson, Professor of Industrial Management and Economics from Chalmers University of Technology in Gothenburg, Sweden. Marcus’s research covers multiple areas of the management and economics of innovation and technology strategy, including governance of innovation ecosystems, platform strategy, and open innovation.

As Scott puts it: “Have you got technology that’s looking for a problem, or a problem we’re trying to apply technology to?

It’s the timeless tension between curiosity-driven discovery and practical problem-solving, and a much-needed reality check in today’s world where there seems to be a tech solution for everything.

Scott and Marcus discuss the need and challenges for matching the right technologies to the right problems, balancing speed with quality, and prioritising human factors that shape meaningful progress.

Together, they explore the interplay between risk, innovation, efficiency and adoption, and how organisations can repeatedly unlock new value safely, be that through AI or other technologies.

This episode was recorded live at the International Council of Academies of Engineering and Technological Sciences Conference and is part of a series in partnership with the Australian Academy of Technological Sciences and Engineering, or ATSE for short.

We hope you enjoy this episode.

+++++

Scott Powell: Well, hello and welcome to Engineering Reimagined. Today I'm joined by Marcus Holgersson, Professor of Industrial Management and Economics at Chalmers University of Technology in Gothenburg, Sweden. Welcome Marcus. It's great to have you here.

Marcus Holgersson: Thank you, my pleasure.

Scott Powell: You have built a career at the intersection of technology, economics and innovation strategy. What first inspired you or attracted you to this area of work?

Marcus Holgersson: So my background is both in engineering and economics, and in the intersection of engineering economics, I think innovation fits very well. And innovation processes are very much about creating new technologies, but also matching those technologies to the markets and to the needs of the customers, understanding that match between technology on one side, market on the other side, and in between there, there are strategies, business models, and so on that need to link those two sides and that's what I find most fascinating.

Scott Powell: So one of your areas of research has been in the healthcare sector. We've seen this industry evolve dramatically in recent years, shifting from really highly centralised hospitals to a more decentralised model using GPs and other medical services. At the same time, the pressure on hospitals during crises, whether they be emergencies like bushfires or floods, has shown a need for a more resilient, distributed system of care. If we combine this reality with the rise of digital platforms, how do we rethink the design of healthcare, both from a physical and a digital sense, to build a system that can support society's health into the future and provide a better patient experience?

Marcus Holgersson: So our interest in healthcare came very much from the platform side and platform strategy from other settings. There is a lot of interest in digital platforms in the strategy field as such. There are certain learnings from these other types of digital platforms, like Amazon, Facebook, Uber, Airbnb, and so on. And then we saw a big increase of digital platforms in healthcare, like remote healthcare in different ways. The problem we've seen is that they've used the same type of strategies as in other settings and mostly failed. So our interest in this field was very much fuelled by trying to understand what is different in healthcare if we compare it to other settings where digital platforms are used and how do they need to adapt their business models and strategies to make digital platforms work in a healthcare setting.

For example, in most other settings, digital platforms build very much on network effects, meaning that the more customers you have, the bigger value you create for other customers. And that's a big engine for most platforms. While in healthcare, we don't see that. It's a completely different value proposition, where typically, more patients mean worse healthcare for other patients, right? So we've tried to figure out how can you design digital platforms in a healthcare setting that actually work? And one main finding is that while in other settings you try to build a big and broad platform to build as much user base as possible, in healthcare, many of the most successful cases have been the niche players, addressing one disease very well, very efficient, and growing from that.

Scott Powell: In that ecosystem then, how have you tackled the challenge of patient confidentiality as you thought about sharing data?

Marcus Holgersson: Data sharing is important in many sectors, but it's maybe most critical and most dangerous, so to say, in healthcare. And there are various strategies you can use to share data, decreasing the risks, like anonymisation, modularisation of datasets and so on, and we see now how different actors develop these strategies in managing the datasets in ways that make data sharing more safe.

Scott Powell: So if we think about that in terms of extending that into intellectual property, and we consider this idea of open innovation, one of our biggest challenges is how much do you share versus how much do you want to hold onto for commercial competitive advantage. So for companies like Aurecon, we've got a lot of intellectual property around timber engineering. And so we're grappling with this idea of how much do we want to share that intellectual property for advancing the industry and particularly the timber engineering industry. So from your experience, how should organisations strike that appropriate balance between protecting their IP but also collaborating in ways that drive industry-wide transformation or industry-wide improvement?

Marcus Holgersson: I think at the core of that is actually creating these assets in the first place so that you can share it in a more controlled way. So, for example, patenting can be one way of doing that. So you have patent rights and then you can coordinate your sharing by using license rights related to those patents. So that's one solution, like 'propertising' a bit of your knowledge base, and by doing that you can govern that sharing in ways that suits your business needs and hopefully the needs of others as well.

Another part of this is also thinking about how you structure that technology. So, modularising a technology, for example, if it's a system technology, thinking about, can we divide these into this set of inventions, for example, into modules and thinking about, okay, what modules do we need to share, maybe completely openly, like open source? Other modules might be better kept secret, others might be patented and so on.

Scott Powell: This idea of balancing risk and quality and human adoption. So in this rush to innovate, the real challenge isn't just about moving fast. We've got to balance making sure that there's a quality in the innovation as well, and particularly in engineering, it's important that we get it right. How do you make sure that you safeguard quality and safety when you're breaking new ground with innovation at pace?

Marcus Holgersson: It is of course difficult to know for sure, because if it's innovation, it is new, right, and you won't know the consequence of that technology. And looking back, it is sad to see that there have been many unintended consequences from many of our new technologies that we are now struggling with instead. We see now over 70 per cent of venture capital investments in the US go into AI. Much without understanding all the consequences, right? And understanding those consequences is very difficult, but I think just trying to do it, being aware of the fact that, okay, every new technology might have unknown externalities or negative consequences, might have unknown positive consequences as well. But just accepting that fact and trying to work with it in whatever ways needed.

Scott Powell: How do you balance speed and freedom, but with appropriate governance so that you can try and at least think about those unintended consequences of what this innovation might play out?

Marcus Holgersson: Well, I think regulations have an important part to play. But we have also the consumer side, the public side of things, and balancing the risks and the regulatory approach to deal with those risks with the private side pushing innovation. But also, as consumers, being aware of these downsides because in the end I think that's most important.

Scott Powell: One of the things that you've emphasised in some of your research is the critical role of human in technology adoption and what cultural or behavioural factors, be they societal or organisational, do you think most influence whether a technology delivers its promised value through that broad-based adoption.

Marcus Holgersson: The main problem then is that we have externalities, meaning these negative consequences outside. So I think as long as technology, the impact of the technology is on the customer and the producer and the innovator, we can deal with those settings fairly well, but as soon as technologies relate to all these other impacts on the global, in the world and our neighbours and so on, then we still have problems. I think one very important area for future research is how can we design systems that better account for these negative consequences?

Scott Powell: So in the exploration of ideas, we've got a framework to consider the consequences of whatever we're coming up with.

Marcus Holgersson: Yeah, like it's not only about technological innovation, I mean, it's very much about some sort of systems innovation here. How can we design systems in a better way, like societal systems that better account for these negative externalities?

Scott Powell: Yeah and then applying those in organisational settings as well should add value to an enterprise. Technology is only as powerful as the people who adopt it. What role do you think leaders play in shaping the way organisations unlock that true organisational or operational efficiency through the benefits that technology can provide?

Marcus Holgersson: I think that's very important because in the end, it's not only about the technology. It's just as much how you match that technology with the strategy and the market needs and so on. And the leaders have the responsibility to guide that matching basically between engineering and markets and business design.

Scott Powell: One of the things I quite often ask is when somebody comes and presents me with a new technology, is talk to me about the problem we're trying to solve. Have you got technology that's looking for a problem, or have you got a problem that we're trying to apply technology? Have you seen that come out in research where, you’ve got to get the balance right, so sometimes technology might solve a problem you don't even know you've got but also, more often than not, you are looking for problems to be solved with technology?

Marcus Holgersson: Absolutely, just take the AI push right now. I mean, it's clearly pouring investments into a technology where we don't really know what it's to be used for. There are use cases, many potential use cases. But much of these investments are just pouring into the hope that they will find some use case in the end. And that's fine, but maybe we need also to be better at addressing the real use cases we know now. Sometimes the technology push into these areas where we don't really see the use cases are because you want to build a very strong position. I think that's what we're seeing now in AI. So, there is a potential here to build, a technological lock into some specific areas, and some firms can take a very strong position there. And those firms will benefit in the long run by holding that position. Working towards more narrower use cases don't have that potential.

On the other hand, you get that direct value creation from something. I think, relates to the healthcare case we discussed before, because these niche actors really find specific, small use case. Which might have a big value for specific subsets of the population, and they address those very well. And sometimes that's enough, I mean, that creates a lot of value. We've seen how those then have grown from there into more general actors addressing other niches as well.

Scott Powell: Yeah, so these niche providers have gone after a specific problem, not tried to go too broad, but solve something where there's a need, a need to respond, and they've seen that be successful. So not trying to do too much, but really just focusing in on a problem and then using technology to solve that. Interesting. So, I guess that is a good segue into this next topic of open innovation. In your research, you talk about AI's ability to enhance or enable or even replace traditional collaboration models. Based on that, how would you advise leaders to advise where to integrate AI into their day-to-day operational processes?

Marcus Holgersson: Innovation requires, typically, collaboration. Innovation might happen within one person as well, but typically it requires multiple people. But AI can sometimes replace a bit of that, because suddenly you have this technology that can provide ideas, feedback and so on into the system. Especially quick feedback. Going back to these use cases addressing actual needs rather than just building a technology, of course, you can use AI to quickly iterate and so on. We've seen that AI is very good at finding the bad ideas. So let's say you crowdsource ideas, right? You have a problem, you want to find different solutions to that problem.

As a big firm, that might be a good idea to do sometimes. We've seen in research that AI can fairly well take out the bad ideas from that set of ideas. At least the studies so far have seen also that AI isn't that good at finding the really top ideas. So if you want to improve, build a new technology, innovate something in your business, you can make that process more efficient by just maybe cutting out the bottom half, the bottom 70 per cent of these ideas to improve things. And focus your efforts on those 30 per cent that really have that potential. That will save resources or improve the progress because you can spend much more time and effort on those really high impact, high potential type of developments and ideas.

Scott Powell: One of the things you talk about is that you see AI changing the way ecosystems are governed, especially when we consider the more decentralised or open-sourced AI models that we have exposure to, and also the onset of new platforms and how they integrate and cooperate. What's your thinking there in terms of what the next chapter might look like?

Marcus Holgersson: The big idea now is these multi-agent systems, right, and how to integrate these different systems in ways, and that means it's fairly difficult to know what's going to happen, but I think there will be challenges relating to that. So, governing those systems of different AI, so who is owning what, and who's owning the output, and so on. But there's a huge potential. Regulation has to follow quickly. Firms have to follow quickly as well.

Scott Powell: In terms of the potential there and the potential that AI presents. Capturing real returns for organisations can be a challenge, right? You identified two key areas in a recent paper around bottom line efficiency gains, taking costs out of your business and then top line growth through AI enabled businesses. Just talk us through the approaches and why complementary assets like data and capabilities are so critical in making these a reality.

Marcus Holgersson: So this is mainly relating to large incumbent actors trying to adopt AI. So we studied closely one such big actor making huge investments in AI, trying to create value and capture share of that value. And we wanted to understand what are the challenges in moving beyond just adopting it, but actually creating value and capturing that value. And throughout that research process, we started to see that, okay, there are two major business models or ways of capturing that value. They use it either for internal efficiency gains, like improving the process internally, which leads to bottom line improvements, hopefully improving profits simply by you know...

Scott Powell: Cutting costs.

Marcus Holgersson: Yeah. And when you do that by improving efficiency, you rely on resources you have internally in your firm because you control all those assets. If we compare it with the top-line growth, building new business, a lot of that business rely on external, complementary resources, external data. Plus, you have to communicate the value of that service somehow. So we identify these challenges in terms of communicating the value that is being created to your customers is not very easy if you're an incumbent actor with established customer relationships. You don't only have to communicate the value internally, you also have to convince your customer to pay for whatever. And that might be possible sometimes if you improve, for example, energy efficiency at your customer site.

That is measurable, and you can probably capture a little bit of that value, but if you improve service quality, if you create new businesses that are uncertain, that is much more difficult. But not only that, you also have to typically access customer data and that comes back to the question of data sharing. Of course, most customers would be concerned to share their data, so you have to find ways of navigating that data sharing space, which is also more tricky than in the internal case of improving your own efficiency. And finally, when you sell businesses based on AI, you also have to have AI capabilities in more parts of your own organisation. Suddenly, you need it on the market side, or the salespeople, the distribution, everyone needs to know a bit at least about AI, while for internal efficiency gain, you basically just have to create the technologies and introduce them in your processes and the rest will work itself out. So it was interesting to see how different it played out and this hopefully can help other organisations also to think about, where can we most productively use AI and maybe in the short run start with those easy use cases and then think about the other cases in the longer run.

Scott Powell: One thing that particularly interests me is this interplay between risk, innovation, efficiency, and adoption, and how organisations can repeatedly unlock new technology-driven value, be that through AI or any other technology. But without introducing those unacceptable risks and then realising the value through adopting these technologies, what advice would you give businesses or enterprises and maybe in a broader sense, society, in terms of being able to unlock and manage that complexity?

Marcus Holgersson: There will be risk with innovation. So, it's about understanding innovation processes. Having management and engineers trained in not only technology, but also innovation. And having capabilities to evaluate the impact on society, the impact on your customers. So it's very much about knowledge, and gaining that knowledge that goes beyond engineering. And we, just this morning here at the conference, talked about T-shaped people. So not only engineers, but they need to have these other capabilities, be it in economics, management, social science, psychology and so on. And having those T-shape people would help to address these risks while staying innovative, while creating that adoption we need to actually make an impact.

Scott Powell: Marcus, it's been great to unpack these topics with you today, we really appreciate you taking the time to share your insights and look forward to seeing what your future research might hold. So, thank you very much.

Marcus Holgersson: Thank you.

+++++

Maria Rampa: I hope you enjoyed this episode of Engineering Reimagined with Aurecon Chief Operating Officer Scott Powell and Marcus Holgersson, Professor of Industrial Management and Economics at Chalmers University of Technology.

If you’d like to hear more fascinating conversations from the world of engineering, science and technology, follow Engineering Reimagined on Apple or Spotify, or visit aurecongroup.com for more information about the podcast and our incredible guests who share their expertise and insights.

Until next time, I’m Maria Rampa and thanks for listening.

Unlocking the value in digital transformation

How can we match the right technologies to the right problems, while balancing speed with quality, and prioritising human factors that shape meaningful progress?

In this episode of Engineering Reimagined, Aurecon's Chief Operating Officer, Scott Powell, speaks with Marcus Holgersson, Professor of Industrial Management and Economics at Chalmers University of Technology in Gothenburg, Sweden. Together, they explore the interplay between risk, innovation, efficiency and adoption, and how organisations can repeatedly unlock new value safely, be that through AI or other technologies.

“The real challenge isn't just about moving fast,” Scott Powell said. “We've got to balance this idea of making sure that there's a quality in the innovation as well, and particularly in engineering, it's important that we get it right. How do you make sure that you safeguard quality and safety when you're breaking new ground with innovation at pace?”

This episode was recorded live at the International Council of Academies of Engineering and Technological Sciences, Inc. (CAETS), conference in Brisbane.