Miika Soininen

Hello, and welcome to Vierema, to the Ponsse factory where we are now. Here you can see, in the background, one of the first serial-produced Ponsse forest machines. We will be here today sharing the story of our journey to data-driven forestry.

My name is Miika Soininen, and I work as Director of Digital Services and IT at Ponsse. I have been working about four years with Ponsse. I joined Ponsse in 2018 as an IT manager responsible for the business software running Ponsse's business. Back then I used to have a three-person team. Since then we have been growing our team, and almost immediately when I started, two of the people in my team quit. We have been going a long road since then. I now have a team of 36 people, plus we have quite a big amount of partners working for us. Our digital services and IT development team is now about 90 people.

Before joining Ponsse, I had experience in different positions, both in software R&D and business IT. I have been working a lot with different kinds of data-driven services and have seen different kinds of development projects during my career. Before joining Ponsse, I was working as managing director of a consultancy company, which had about 100 people working. Now I report to Group CEO Juho Nummela and have also been a management team member since December 2020. Here I also have with me co-presenting, Nikhil Sharma.

Nikhil Sharma

Thank you, Miika. Hi. My name is Nikhil Sharma. I am working as Director of Product Management at Copado. To take a step back and look at my career: I studied mechanical engineering and worked quite many years researching a lot of mechanical-engineering-related projects, only to realize that software is really getting very crucial to machines. It is becoming the heart of every machinery. That is when I gradually started moving toward computer sciences and IT.

For the last eight to nine years, I have been working in testing and test automation, essentially because I like to break things and prove things wrong. That is when I also met Miika. I was working as a test automation engineer, worked for quite many companies, ramped up their test automation projects, and then in the latter half I have been working in the capacity of product manager for a product called Copado Robotic Testing. We will talk about it later in the slides.

Copado is the number-one low-code DevOps and testing company enabling enterprises to reach what we call DevOps Nirvana. I am very much focused on the testing side of the business, so that will be the main focus of my talk track.

Miika, when I look at harvesters, and many other people look at harvesters in the forest somewhere logging wood, people do not think about software. Where is software present when there is a logging harvester in the forest?

Miika Soininen

Nowadays, basically, software is driving the harvesters. It is totally a software-controlled system, and data is now really essential for our customers. In fact, this is one of my favorite quotes from a customer. When I heard this last June from the customer, I had to write it down immediately. Basically, our customer was saying that they do not want to operate the machines anymore if they do not have the data from the machines. That tells us there has also been a big change in the market and in what software means for our customers: they need the data so much that they do not want to run the machines anymore and harvest the wood without the data.

To put it into context, this particular customer is harvesting in Canada. Their conditions are good to understand because they are operating over quite a large area. It might be 200 miles from the workshop to the working areas, and it might be an eight-hour drive there. They want to know exactly how the harvesting is proceeding, what the production of the harvesters is, when to send the forwarders to pick up the logs from the forest to the roadside, when to send the trucks to take the wood from the roadside to the mills, and also how the machines are working, whether there are any fault codes, how much fuel there is, and so on. It is really important for them to make a profitable business that they can have a good understanding of the operations.

Nikhil Sharma and Miika Soininen

Nikhil Sharma: So we are essentially talking about a lot of data and a lot of perspective to that data.

Miika Soininen: Yes. There are many stakeholders utilizing the data, and the data is blended there for sure.

Miika Soininen

To open up a bit what we are discussing, maybe a few words about Ponsse as a company. We are the leading forest machine manufacturer and concentrate on cut-to-length method forest machines and services supporting the machines. The essential part here to understand is cut-to-length. There are still a lot of harvesting operations done manually with chainsaws. Then there is a full-tree method, where the trees are taken as a whole from the forest to the mills, and it requires a lot of equipment and machinery to take the wood from the forest to the mill like that.

In the cut-to-length method, we are getting the orders from the mill side, and we are just taking what is required to fulfill the order. It is driven by data, and we have accurate data from the forest machines that can then be utilized throughout the chain. Actually, the forest machine is really a high-accuracy measurement device loaded with software.

To tell you a bit about Ponsse and give you the scale: we are operating on all continents in the world, in over 40 countries nowadays. The main markets are basically where forest is naturally growing, in Northern Europe, Central Europe, and Northern and Latin America as well. There are over 2,000 Ponsse employees globally working to support our customers. Our turnover last year was 750 million euros.

Here are a few pictures from our factory here at Vierema. Here at Vierema we manufacture all of our products: harvesters for cutting the trees, forwarders for the local transport of the wood to the roadside, and also the loose harvester heads. We have here the most modern forest machine factory in the world. Something good to understand about Ponsse is that we do not want to be the biggest in the market, but we want to be the best at what we are doing. That is one of the reasons why we have the most modern forest machine factory here.

To give you a bit of an overview of our operations, I want to highlight a few things. One of the most important things for Ponsse is customers, and that is visible basically in everything we are doing. It is not something that we just put on slides, but it is in the heart of everything we are doing and visible in the daily work with us and in communication and cooperation with our customers. What we often hear from customers is that it is different to work with Ponsse. People are really down to earth, it is really easy to get access to Ponsse people, and Ponsse is really willing to implement what customers have asked.

Our vision is to be the preferred partner in responsible forestry. We want to help our customers and our customers' customers succeed in their business and take care of the forest responsibly in the future. One of the key changes we have made during the past couple of years in our strategy is that we have shifted the focus from the product offering to the solution offering, meaning that we want to understand customer needs and fulfill the customer needs with our solution, not only with the machines.

Nikhil Sharma and Miika Soininen

Nikhil Sharma: That is a very interesting point. You are operating in all the countries you mentioned, and the scale is large. The ecosystem is huge. What does it mean to move from a product approach to a more solution-oriented approach? What kind of complexities and challenges does it bring to Ponsse, and especially to the digital services part of it?

Miika Soininen: Of course, it changes the expectations from the customers, and this is also the request from the customers that we are getting: they want to utilize the data. But it also means that we are taking different kinds of responsibilities and different kinds of deliverables. For example, for the machine and the service, we do not just do the maintenance of the machine; we promise, for example, a certain uptime for the machine, how the machine is operating.

What we have raised here in the center is digital services tying everything together: training services, lifecycle services to take care of the machine throughout the lifecycle, consultancy of the operations, and also training, where we can utilize the data and focus on the training as well.

Nikhil Sharma: So it is essentially moving more toward harvesters as a service. It is really like a service model.

Miika Soininen: Yes, exactly. Moving more toward value-added services.

Miika Soininen

To open up a bit about the operational environment, what we have here, and what Nikhil was also referring to, it is good to understand how forestry works altogether. Basically, our customers' forest and harvesting operations are at the heart. The forest machines are in an interesting position within the forest because there is a lot of information coming from the forest machines that can be utilized for steering different parts of this ecosystem.

For example, if forests are thinned, so only part of the trees are taken from the forest, the information gathered from each log cut can be utilized to plan forest management because it is really accurate information about forest resources, and growth models can be built by utilizing information coming from the harvesters.

Then one of the big problems for different stakeholders is transportation. There needs to be information: how much harvested wood there is, what kinds of assortments, and where, so that transportation can be optimized. If you consider harvesting, for example, in Northern Europe, the work areas are really small and the operations are also smaller, so there are usually a few people and a few machines working on the site. But it requires a lot of planning to collect all the wood from the different work sites.

On the other extreme, we have large customers where it is more like agriculture, for example eucalyptus plantations where all the trees are planted and a huge amount of wood is coming. The operations are so big that there can be thousands of loads of wood per day coming from the forest to the mill to keep the mills running. It is a really big data-driven challenge for us and for our customers to run everything here, and also to address all the different customer needs, because all the customers are equally important for Ponsse.

Nikhil Sharma and Miika Soininen

Nikhil Sharma: It is very exciting to see the sheer complexity of this ecosystem. It must be bringing you a lot of challenges, doesn't it?

Miika Soininen: Yes. It brings challenges for us. When we started a few years back, one of the big challenges for us in meeting customer demands was the scattered architecture we had back then. That, of course, leads to it being quite slow to build up new solutions. Like I described in the beginning, we had to build the organization to be able to deliver new kinds of digital services and also set up the processes that support delivery of the digital services and maintenance of the digital services. That is related to the scattered architecture: the organization was also a bit scattered, and the processes were not harmonized, leading even more to scattered architecture and maintenance problems.

One challenge I want to highlight here is connectivity. If you look at the picture in the background, you can imagine that we are not working in city centers. Our machines are operating in areas where there is not good coverage. Our challenge is to create real-time data-driven solutions for areas where there is no network connectivity. That is one of the big challenges we have had to overcome.

Nikhil Sharma: It would be quite a sight to see you working in city centers. To talk about these challenges, and that is very typical of what I also see in many other customers, speaking of these traditional industries and other manufacturing industries: they have been there for so long, and they have lived through all these different stages of software development. You have seen mainframe, cloud applications, on-premise applications, different platforms, waterfall ways of working, agile ways of working, going DevOps.

You see this typical architecture, which is a very complex IT landscape, where different generations of applications talk to each other through homegrown APIs with different release cycles. Some of them might have a release cycle of three weeks, others three years, and for others maybe never. The end customer does not see the complexity of this IT landscape. That brings a lot of responsibility onto the shoulders of people like you: how do you cut that complexity out, make sure that things work interoperably, and the end customer gets confidence and can work seamlessly with your services?

Miika Soininen

That is exactly the challenge we have been tackling. To open up a bit how we have addressed the challenges that you described really well: we started basically almost immediately when I joined Ponsse and started working with one customer demand, the demand for data. What we did back then is that we actually built from the ground up within six weeks from signing the agreement to implement the new platform, and then we delivered the new APIs for the customers. That included setting up the new environments, building up the application to handle the data, building up user access management, and everything.

This was one of the first success stories already in 2018. It showed us that it is important not to miss the customer demand, but to deliver fast. We understood already back then that we actually created quite a lot of technical debt and took a lot of shortcuts. But we also learned that when we can deliver fast, it is easier to build on top of the deliveries than to build the perfect world and see that the customer demand has already gone and someone has already solved the problem.

Since then, we started systematically building toward the situation where we can be the leader of data-driven forestry. One of the key elements was that we created a shared vision: where we want to be and how to get there. One of the key aspects has been that we have had a lot of trust at Ponsse and good cooperation with the management team and different stakeholders. They have been willing to invest in the digitalization and digital platform that we have built.

Basically, we have built a lot of new capabilities, including, for example, the cloud-based data platform to hold all the data, not only coming from forest machines, but also from customer processes and Ponsse business processes. We have developed API-led architecture. We have developed a lot of different components and are trying to build the Lego blocks to meet customer demands. At the same time, we have developed our processes and tooling so that we can learn to deliver faster.

What we have accomplished is that we now have, as we speak, something that brings a lot of value for our customers. We have a new value-added service offering built on top of the foundation that we have done, and we can truly provide something unique for our customers.

Nikhil Sharma and Miika Soininen

Nikhil Sharma: You pointed out in the slides that learning to deliver fast is essentially a very big learning for a company, or for any company out there, to build that competitive edge and also sustain that competitive edge over time. It is very close to the Lean foundations: fail fast and deliver fast. You can judge the quality of a product by thinking of how small a feature you can put out into production, how fast you can do it, and how sustainably you can do it over time. That is a great learning.

Miika Soininen: I agree. What I really believe is that there is not any customer value before we can deliver something. When we can deliver new features, then we can at the same time get faster feedback from customers. When we do frequent deliveries more often, we can also improve quality. That has been one of the driving factors for us, taking quality assurance metrics also as key decision-making metrics for us.

Nikhil Sharma: I would like to take a step back and portray this picture for our audience. You probably have been many times in a situation when there has been a software update and it led to degraded software performance. That leaves a lot of frustration: why did I upgrade in the first place? But maybe you are not in the middle of the forest when you do it.

If you think from a logger's point of view, think about a forest that is far, far away with very low connectivity. You are sitting inside this harvester doing your job, and everything has been working fine. Then we added more features to improve your work or data collection, and all of a sudden it does not work. How would you feel as a logger sitting in the middle of the forest, not knowing about all the complexity that Miika and I have been talking about? Perhaps very frustrated. That is when users start to lose trust and faith in a company, and that is something a company cannot afford.

The way to solve it is: how do you increase customer value through these quicker feedback loops? How do you bring high automation to your upstream and downstream of software development, bringing it into different phases of testing, release management, continuous integration, and continuous delivery, so that you increase quality, with fewer bugs leaking out to production, and build confidence not only in the developer teams and stakeholders, but also in that logger who is in the forest? So our logger is a happy camper, or a happy logger in this case.

Miika Soininen: Exactly. It is really important to balance making everything work and, on the other hand, competing more and more on software features. We need to be able to bring new features even though we have a really complex environment. Managing this is something we need to learn to do even better in the future as well. A high automation rate is one of the key solutions to that for sure.

Nikhil Sharma: That is essentially where Copado Robotic Testing came into the picture when it comes to test automation. Why did you pick Copado Robotic Testing?

Miika Soininen: In software development altogether, we do not develop our own tooling for software development. We picked Copado Robotic Testing so that we could get started really fast and build confidence just to get those releases out and make sure that we can release more often and the releases are working.

Nikhil Sharma

To give a brief about Copado Robotic Testing: companies like Ponsse and others, where you have this complex IT landscape, have very taxing demands from a test automation product. You need a truly extensible product to address all of those needs. That is what we set out to build.

What we saw in the market when we started building this product in early 2016 was that products out there did not talk about software development and testing as an integrated part. Software development was moving to cloud. Testing was still a laggard. It was still a tool on a desktop with some tester sitting in the corner of the room testing and trying to put that information back to ALMs.

Also, people did not talk about test automation infrastructure. That is where your test cases run, and it is such a big investment that if you do not take care of it, it starts cannibalizing your test automation efforts. Lastly, but importantly, the tools of history were quite limited when it came to extension. If you needed to build an integration or extend it, it was a big system integration project.

So we built Copado Robotic Testing as a cloud-native test automation, truly extensible product. It uses Robot Framework's execution engine, and that gives you the capability to extend. We provide a lot of standard libraries and a consistent way of doing test automation. It is multi-cloud and multi-platform, be it web, mobile, or on-premise applications. For example, Ponsse did a remarkable job: they extended the product to test their APIs. That is something they built themselves, and it works in the product. All of that means you can run all these different test automation initiatives and have the results in one place behind a URL accessible from anywhere on this planet where you have internet. That is how Copado Robotic Testing works.

How did we start working together, Miika? It was in 2017, wasn't it?

Miika Soininen and Nikhil Sharma

Miika Soininen: The cooperation started already before I joined Ponsse in 2017, and we have been expanding ever since and utilizing it more and more over time. We started with Ponsse Manager, which is the brand where we offer all the digital services for our customers, and we have been building automation for the web and mobile applications there. Since then, we have also expanded, for example, to our business-supporting systems, ERP system automation, and, like you mentioned, API testing.

Over time, we have learned to rely on the information we are getting from testing and also the DevOps pipeline so that we know the situation of the releases, and we can make decisions based on the information we are getting. That gives confidence for the developers and also for the customers and end users of the system. We are also saving quite a lot of money compared to the situation before, when we used to do all the testing manually. Basically, nowadays, we could not anymore cover all the testing needs manually. We must have automation to support us in the operations.

Nikhil Sharma: I have to really say thanks. It has been a very exciting journey cooperating with Ponsse. It is a really passionate team. They know a lot, and they have greatly shaped the product. It has not been just us providing the product. It has been great bidirectional cooperation where all the valuable feedback that we have received from Ponsse has largely shaped Copado Robotic Testing as a product. A great thanks to you and to the team for all the cooperation so far, and it will continue.

Miika Soininen: For sure. Like in our own work, we want only to be the best, and we expect the same from the partners as well.

To summarize the presentation, I would say there is a big demand in forestry altogether for data-driven services, and we are seeing that demand as we speak. That has put us on the journey to evolve Ponsse to have, in addition to great products, also great services. We have been working hard within Ponsse, but especially together with different stakeholders and different partners, and a lot with our customers, listening to customer demands and trying new things with our customers. To succeed in the tight competition in the future, we see that we need to work together.

Like Einari Vidgren, who founded Ponsse in 1970, used to say, "It is important for us to do what we are promising and to listen to customers." That is one of the takeaways I want to raise here as well for everyone working, no matter what business you are working in: listen to your customers, and you can walk a similar journey to the one we have walked with the digital offering to be the leader in data-driven forestry.

I really see that data-driven forestry is a big ecosystem. There are a lot of stakeholders that can both provide data in the ecosystem and utilize the data, and there are still countless opportunities to improve and provide value for the different stakeholders, to provide new solutions. What we are looking for here is new connections, possible partners to develop the forestry ecosystems in the future. Also, we are still expanding our great team, and we are looking for new talents to join us.

Thank you very much, Nikhil, for being here with me to present this, and thank you all for listening.

Nikhil Sharma: Same here. The pleasure is mine. Thank you, everyone. It was great learning about the forest industry for me as well. Bye.