Working together on innovation in offshore wind to provide green energy for the future! Kunlabora very much enjoyed working on this Launcher project for Siemens Gamesa together with ARGON Measuring Solutions and Hexagon Manufacturing Intelligence. Working on real stuff like wind turbines is really exciting!
Some time ago in a country nearby a team of black ninjas was working on a huge innovation project for their Cuxhaven factory where they produce some of the world's best offshore wind turbines. For your understanding it might be good to know that the black ninjas are a superteam of Siemens Gamesa working on huge innovation projects. Did you know that wind turbine installations are expected to grow from a 20 GW installed base in Europe in 2020 to 450 GW by 2050? To support this a lot of research and development is done not only to create wind turbines with a higher capacity, but to produce them faster too. That's also why Siemens Gamesa built this modern wind turbine factory in Cuxhaven in 2018, offering jobs to over 800 people.
So, after they had been studying for quite a while, testing with a prototype to support their case, the black ninjas decided the impact of this innovation was so huge they couldn't wait to proceed to the next step. This next step would include the engineering and installation of the fastest and most precise measurement technology. However, to be complete it would need to be integrated in the assembly process to make sure the factory operators could use the technology in their daily work with only a minimal support of measurement engineers. A project was launched to develop a first version of the solution to use in production for a while to get an even better idea on the possibilities for further improvement. This is exactly what we (Kunlabora) call a Launcher project. A Launcher project is meant to run in production early, knowing it might not be complete and perfect, but good enough to start and to learn. It will be changed and improved later. Focus on quality, e.g. providing good and enough automated tests, is very important. Delivering nice-to-have functionalities is postponed until more knowledge is available on the business value they will bring.
For this next step the black ninjas first contacted Hexagon Manufacturing Intelligence, a world leader in measurement technology. They suggested to cooperate with ARGON Measuring Solutions, a niche player from Leuven, specialized in integrating measurement technology solutions on the shopfloor (their references include other big companies like Airbus and Audi). The proposed solution would make use of unique measuring technology developed and delivered by Hexagon Manufacturing Intelligence in a shopfloor setup engineered by ARGON. Of course providing good measuring technology alone is not enough to come to a workable solution in an industrial production environment. To support operators in their daily job on the shopfloor the use of the measurement technology should be as easy to use as a coffee-machine, according to ARGON's concept of the Coffee Machine Principle. And on top of that, one of the important innovations the black ninjas wanted to introduce was an automated calculation of the optimal adjustment based on these accurate measurements. Given the impact and challenge of this project ARGON asked Kunlabora, a company based in Leuven and specialized in building tailor made software following a disciplined agile approach, to cooperate on this task. For building easy-to-use interfaces and incorporating some intelligence in this user interface ARGON and Kunlabora could help each other. This collaboration has proven to be very valuable and enriching, bringing together ARGON's overall project vision and measurement technology expertise with Kunlabora's software development expertise and disciplined agile project approach based on trust and transparancy. This is how two small niche players from Leuven, together make a difference for the world leader in the offshore wind sector.
For a start, we are convinced it is important to get a clear overview of a project. Since in this case nobody had a clear idea of what the final solution would look like, setting up a Service Blueprint phase was a first step. We invited the most important stakeholders, c.q. the black ninjas and our friends from ARGON, with some of their colleagues from Hexagon, for a few intense days of collaborating. As a result of the workshops we all got a better idea of what features and functionalities were needed in the software and when they would be needed. We started with an overview of the impact map. We drew a context diagram and identified users and systems involved. We used story mapping to help us getting a common understanding on what to build, covering not just the success scenario but talking about what could go wrong as well. We did a risk storming session, discussed the process flow and learned about the desired calculations. We went through the story map again to define priorities. By the end of our 3 days together we learned a lot about the project, although we still had a lot of questions. But it seemed good enough to make an estimate and propose a more detailed planning in the following days. On top of that, we got to know each other which made everybody feel more comfortable to trust one another.
With the input of the service blueprint workshops Kunlabora and ARGON further detailed the software architecture in container and component diagrams. With all this knowledge we could make an estimate of the needed effort to develop the software functionalities. We could tell how much time we needed, what kind of team was needed and which budget had to be provided. We bundled everything in a Service Blueprint report to keep track of our agreements. This gave the black ninjas enough information to decide on a further cooperation for this project.
We set up a mixed team with Kunlabora and ARGON colleagues and started building the software in 2 week sprints, planning the stories based on the priorities we defined together and making sure stories with more uncertainties or risk were taken up early. At the end of each sprint we prepared a demo of the developed stories to be able not only to show our progress but to gather feedback early as well. After a few sprints the black ninjas visited us again. This allowed us not only to gather feedback on what had been developed already but also on what we were planning to do for the following stories and features. With a demo of the first stories in mind and the knowledge on what was on the planning for the coming sprints, we could go much more in detail and discuss edge cases and special scenarios. We used some process diagrams, screen mockups and screen flows, and of course our story map to support our discussions. After discussing our idea on how to set up the algorithm we also started implementing that and we soon could give some first feedback on calculation time needed. Since we had to do quite some hardware integrations with the measurement technology and a PLC (Programmable Logic Controller), we started working on the most difficult integrations early to be able to learn whether we would be able to succeed and of course to test the integrations one by one as soon as possible.
Since we were spread over different locations, the black ninjas in Germany and Denmark, ARGON in Haasrode and Kunlabora in Leuven we used Miro for discussing the process flow and screen designs, and tools like Dropbox and Onedrive to share documents. Slack helped us for small discussions, and as the Corona virus COVID-19 came by, we became good friends with Google Hangouts, Skype and Microsoft Teams too. Our disciplined way of working with a daily standup and a sprint burndown on a whiteboard was easily transfered to working as a remote team at the start of the Corona lockdown. We didn't notice any slowing down at all. All these tools helped us to communicate easily and stay tuned during the complete software development phase.
We continued building the software in 2 week sprints, demoing the newly developed stories at the end of each sprint. We made sure to organize demos regularly with the black ninjas too, since they will have to use the software and approve it. We helped them testing the algorithm which is one of the most important parts of the software and as a result we did some extra improvements. In between we also took time to test the hardware integrations (including all measurement technology), even though the final setup of the hardware was not ready yet. We delivered the Launcher version of the software with all agreed-on functionalities, knowing that some nice-to-haves had not been implemented yet. The complete solution was tested using the real hardware and measurement technology in a test setup before being installed in the real production environment. Using the solution in production for a while, will give the opportunity to learn what changes and improvements might be needed before extending the solution or rolling it out in other factories. After a few months running in production the black ninjas and our friends ARGON and Hexagon will certainly have an even better idea on how to improve and optimize the solution. How much software changes will be needed, will be checked then too. What exactly will be the next steps will largely depend on the results of this first real-life testing period.
For now, we can just do thumbs up and hope for a lot of succesful assemblies of the windturbine parts in the Cuxhaven factory in Germany using the new Launcher solution.
We are always happy to explore your needs. Contact us to make an appointment.
Would you like to know more about us? Do you want to try a new idea with us?Contact us