IoT Solutions World Congress, Barcelona
As I’m asking Andrei Ciobotar, relayr’s chief technology officer (CTO) to rewind to 2014 and the start of the company, I’m struck by how far and how quickly the Industrial Internet of Things (IIoT) has come, says IoT Now’s Jeremy Cowan. Even just eight years ago companies like this didn’t exist. There were diverse departments within a few giant enterprises, but mostly they weren’t talking. Hey, sometimes they didn’t even know one another!
Fast forward from its founding and relayr now numbers about 300 people, mostly in Germany, USA and Poland. But it’s what they do that emphasises IIoT’s evolution; today relayr delivers complete solutions for a risk-free, digital transformation and it is supporting customers moving towards Equipment as a Service (EaaS). The company enables industrial businesses to shift from capex to opex-based offerings by combining IIoT technology and solutions with critical business services. So, customers including manufacturers, network operators, and industrial equipment service companies have a single source of IoT financial services or relayr’s own bespoke transformation de-risking instruments. Whatever their starting level of IIoT sophistication, customers can deploy interoperable IIoT solutions that enable their digital and Equipment-as-a-Service transformation.
Jeremy Cowan: Relayr may not be known to all of our readers. What has the company been doing since it was set up?
Andrei Ciobotar: It was started in 2014 and relayr has gone through quite a few transformations. Back then the company focused on more consumer-driven use cases. In 2016 we made the switch to heavier, industrial-driven use cases. The business of IoT is, in general, not very tied to the flavour of asset that you’re connecting to, but you do have to think about an entirely new user experience for industrial customers. And you also have to think about new capabilities that are very relevant in the industrial space.
That led to two acquisitions that are relevant. In 2017 relayr acquired Proximetry, an established name in Poland in IoT, with 10 years-plus experience in edge management and edge devices. So that brought on capabilities to manage gateways and then to roll out software updates to the edge.
And the other acquisition was a company I had co-founded, Neokami, which had machine learning capabilities, infrastructure to operationalise artificial intelligence (AI) models, and obviously the know-how to develop these capabilities. Neokami means ‘new spirit’ in Japanese. That also was a pretty big shift for relayr culturally, because it had graduated from being a Berlin-based establishment to being a distributed company. With Proximetry we opened a new office in Katowice, Poland and with Neokami we opened the Munich office. So, it was a new stack, new capabilities and also new culture, moving to a distributed set-up was quite a change for the organisation to absorb.
AC: As you know, when you dabble in IoT there’s a huge amount of potential use cases and applications, and they tend to be asset-specific. Connectivity and collecting data may be a little more asset-agnostic. But once you drill into the user experience and the insights that drive decisions and value, then it gets really asset-specific. That made us focus a little bit harder on product lines, and gave birth to SKYLER, the product focused on predictive maintenance and remote monitoring for rotating equipment.
Another product is Franz, a remote monitoring tool for elevators. And Equipment-as-a-Service (EaaS) is also something that that relayr is working on. You can think of Franz, SKYLER and Equipment-as-a-Service all sharing the same common IoT foundation, but also introducing new capabilities. And we’re uniquely positioned as part of the Munich Re Group to bring financial structuring and risk services into play. (relayr was bought by Munich Re in 2018.)
Part of the proverbial ‘secret sauce’ in our EaaS stack is this ability to bring expertise together. And to offer a simplified experience to the user base, because, of course, you can’t go shopping for an IoT platform and go shopping for an ERP system and glue everything together. It tends to be a bit of a mess, and you have to look for partners that don’t necessarily click together like they do within the Munich Re Group.
JC: Was that one of the key moments when you shifted your structure to have a financial services company behind you?
AC: What was really interesting is that Munich Re was and continues to look at us as an innovation arm. So, in some sense, relayr has actually carried on business as usual with the same speed, same aggressiveness, just with, obviously, more structure. And, as you mentioned, with the capabilities Munich Re brings to the table specifically around risks, and services, or financial structuring.
That was a huge opportunity for us to bring departments that normally are very unlikely to work together into creating this comprehensive Equipment-as-a-Service package. One of our colleagues in Munich Re very aptly called it ‘Simplicity is complexity resolved’.
JC: Why do you think so many companies fail to understand their customers’ pain points?
AC: I think there’s multiple reasons, one is language. When you have IoT folks sitting together with business folks at the table, it’s very difficult to find a common vocabulary and make sure that the same message comes across. Ultimately, if you’re a software company selling to customers in the industrial space then that problem is only exacerbated. You need to have one foot in the software world, but definitely one foot in the production.
Also, digitisation is a big topic these days, and Industry 4.0, and cyber physical systems, and then there’s a lot of buzz around all of these activities. Sometimes there is a tendency to put a solution in place that looks for a problem rather than the other way around, and maybe be prescriptive with your partners about what kind of solution you would like to have. That may not necessarily address the root cause of your challenges or address the fundamental needs of your business if it’s on a transformation journey. That lies partially on an unrealistic expectation on the customer side, but it also sits with the companies like relayr, IoT companies and service providers have to guide the customer. It’s a journey, and that’s not always in the foreground. A transformation, especially in this business, is more of a joint partnership and a journey rather than a pure software services play where you throw a SaaS platform over the fence and hope that something good comes out of it.
Naturally, there’s also a lot of unrealistic expectations in the market. And sometimes the unfortunate truth is that some of the outcomes that are expected cannot be delivered either by lack of data, lack of tooling, lack of process, any number of reasons.
JC: I think sometimes in IoT we overlook what industrial businesses are actually looking for. Is that fair?
AC: Industry 4.0 and digitisation are really at the forefront of these transformation journeys in the industrial space. And you see these diagrams describing the different facets of what the industry 4.0 set-up looks like and you look at big data and IoT, at machine learning and cyber-physical systems, augmented reality – there’s an entire hat of magic words that you can pull on the topic. But what is maybe not so obvious is that all of these concepts are intertwined. Maybe IoT is not the kind of differentiator that it was a decade ago, the ability to connect assets today to the cloud infrastructure. But I would argue that IoT is really still the barrier of entry and the main stepping stone to get to the rest of these capabilities. And when we talk about big data and machine learning, we cannot deliver any insights that are data-driven without an infrastructure to collect data.
When you’re talking about cyber-physical systems, and we have COVID these days, there’s a lot of buzz in the market around how do we make collaboration easier for remote employees, and also another factory line. How do we interact with the cyber-physical systems? It still boils down to how do I create digital twins of my IoT devices? How do I feed them with data? How do I create a digital representation of these assets? And these are fundamentally IoT problems to solve.
So, in some sense, any business that wants to get to more sophisticated use cases does have to go through this whole connectivity journey first.
JC: Can you give examples of the pain points experienced by your customers? I know customers are wide-ranging but sometimes the problems cross barriers.
AC: Well, I think there’s, there’s the usual suspects. Predictive maintenance is something that everyone’s thinking about. But predictive maintenance is also an ethical, interesting topic to think about more deeply, because when you want to predict if an asset will go offline, for whatever reason, unplanned downtime, you do have to calibrate your system to recognise the behaviours that lead to that time. Most of the time, our customers and partners don’t have this data foundation to even get such a system in place. So, a pain point is not only the need to reduce unplanned downtime, but it’s also the needs to create this foundation of data, which I would argue is even more fundamental. Other customers have had a data collection strategy in place for years but have not had the chance to interact with the data in any way. And that leads to, first of all, unrealistic expectations to what can be achieved with the data.
Secondly, it leads to quite a lot of frustration when the data that’s been collected for years cannot be used in meaningful ways for the use cases that had that they had imagined when they started the collection process. These are all topics that are intertwined. Other topics are operating efficiency, obviously, how do you benchmark your asset against other similar assets and field? How do you turn the knobs, proverbially speaking, to get the performance out of the asset? These are really the most common ones we see out there.
But then, there’s also a fair number of customers that already have some sort of infrastructure in place, they have a reasonably high digital maturity, or a reasonably high sophistication already in their infrastructure, data collection strategy, or even capabilities on the machine learning side. For these customers, you have to go a little bit higher up. They start thinking about topics around improving efficiency of a process as an example. You’re looking at fingerprinting data for liability purposes, it gets a little bit more sophisticated.
JC: If you’re receiving data coming from your customers’ machinery, and you’re analysing for predictive maintenance purposes, you’re engaging at such a high degree within their company and their data. How many companies you’re competing with have similar levels of engagement?
AC: I think that depends on the business model. When we look at use cases around SKYLER as an example, looking at issues with the rotating equipment based on vibration data as an example. It goes without saying that the data we capture belongs to the customer. So, in some sense relayr is a steward to the data.
If you look at this model, there’s quite a few companies that operate from with a similar principle. I think this model is quite ubiquitous these days. But when you look at a set-up like Equipment-as-a-service, things get a little bit more interesting, because many times you’re buying the assets, which essentially makes the data yours. But at the same time, this asset is deployed to an end customer site that may or may not be comfortable with the OEM or relayr, being exposed to their production data. Sometimes, maybe it’s even a trade secret. When I look at a bottling plant, Coca-Cola is an example. I think that’s where we relayr and Munich Re are in a more privileged position in the market and can be trusted to be stewards of the data. In my opinion there’s very few actors out there that would enjoy this level of trust.
JC: You’ve talked about the shift from capex- to opex-based IIoT services. Could you expand on that?
AC: It’s lowering the barrier of entry for acquiring equipment. It’s the ability to offer services, additional services across the lifetime of that asset, bundle the asset for data-driven services. And capture more of the value chain ultimately. And I think that’s the fundamental part of it. But it comes with a lot of side effects.
If you look beyond what it means, on a balance sheet, and you look at the side effects of this capex- opex transformation, it brings with it quite a few interesting challenges to solve. One is that the focus moves to the performance of the asset. It has an intrinsic need for remote monitoring and IoT solutions, having a fleet of assets in the field that you potentially own or liable for? I think that’s a no-brainer for IoT systems. It really boils down to the partners capturing more of the value creation chain.
Comment on this article below or via Twitter: @IoTNow_
