Thanks to SST and CQM's solution, Tennet is able to put the pieces of its planning puzzle together in a smarter way: ‘More time to promote the energy transition’
The energy transition is in full swing. As such, Tennet, the grid operator that supplies power to over 41 million end users, has even more work to do these days. Such as all of the newly constructed wind and solar energy farms that need to be connected to the electricity network. Efficient planning is therefore absolutely essential. SST Software and datascience specialist CQM developed the solution together: a work package planner that makes putting the pieces of the planning puzzle together faster and easier.
make your mark with technology
Solar energy farms, hydrogen production, energy storage, electric cars: developments that are accelerating the energy transition are taking place in quick succession. Some of that newly generated energy needs to be fed into the grid. Fortunately, Tennet, the operator of the high-voltage network in the Netherlands and large parts of Germany, is facilitating this process expertly. This does, however, mean that the company's activities, which also include the installation and maintenance of high-voltage lines, have increased exponentially.
‘You need to look at it like this: we have 300 percent work to do, of which we can carry out 100 percent', says Pascal de Groot, member of Tennet's Integral Planning Team. ‘In fact, we can really only manage 80 percent, because we're not only working on new projects, but also resolving faults on the existing high-voltage network. We need to schedule time for that too. So it is not just a question of scheduling all that work, but also of setting priorities and making choices. Coordinating all of those activities is a complex business.’
In short: Tennet faces major planning challenges. Especially when you consider that the public company is carrying out these operations, no fewer than 300 ongoing operations in total, with approximately 6000 people. De Groot: ‘Our work used to be more transparent. Now it has increased exponentially and we have to deal with various interests. Work can no longer be planned the normal way.’
That normal way used to be manually inputting data into software packages. ‘We ran into limitations,' says De Groot. ‘I graduated in Knowledge Technology and wondered: could we not further automate the decisions that planners have to make?’
the ideal combination
That's when the quest for a solution started. De Groot found a listening ear in the University of Twente and tech companies. One of them was SST Software. One of SST's employees, Kamiel Niezink, had previous experience in Tennet's line of work and had even collaborated with them. ‘So I was already familiar with the challenges faced by Tennet,' explains Niezink. ‘At the time, I was amazed at the process of compiling that intricate puzzle. I saw people toiling for days to complete the schedule. Based on my knowledge of software, I thought: that process could be smarter, easier, and thus automated to a greater extent.’
Another tech company had exactly the same thought: CQM, a fellowship of mathematically oriented professionals that helps organisations gain insight into complex processes. Niezink: ‘CQM thought and spoke to Tennet in the same way. That's why Tennet put us in touch with each other. Our worlds collided, and it proved to be the ideal combination. I immediately thought: a good solution could come out of this. CQM is more involved in the mathematical modelling behind it, whereas we build the software. One plus one equals three in this case.’
De Groot, from Tennet, also realised this quickly. ‘Our problem is so complex that we were going to need various specialists. Specialists like SST and CQM. They could both bring something to the table, so we asked them to combine their services into a total package. Luckily, in the end that proved to be the best approach. The advantage was that Kamiel has project management experience in our line of work. He knows what working on a high-voltage network entails and understands the importance of periods when the voltage is switched off; that those are very difficult to schedule because it needs to happen in phases.’
SST colleague Menno van der Werff (consultancy and design) and Jacob Jan Paulus (algorithmics, modelling) of CQM complemented the project team that was put together. ‘I am responsible for the calculations, to make sure that the outcomes are correct,' says Paulus. ‘SST is responsible for embedding into the workflow, for the user-friendliness of the product. What is great about our partnership is that we gave each other freedom in terms of the grey areas.’
Niezink: ’Our work focuses on how end users interact with the solution, but CQM is responsible for all of the smart logic. We gather the data, present it to CQM, and are confident that they'll use it correctly.’
The issue was clear, as was the division of roles to achieve a solution, but the outcome of our partnership was uncertain when we started. ‘It's like developing the first hydrogen car: it doesn't exist yet, so you have to invent it. Granted, we had a smart, software-based work package planner in mind, but we did not know exactly what it would look like. We established that throughout the project, in sessions geared towards exploring the needs of the client. Very agile. What should the logic be? How can we gather that data? What are the priorities? And what are the variables that need to be taken into account?’
The sessions took place digitally because of Covid. ‘I have never met Pascal face to face, but it doesn't feel that way,' laughs Niezink. Based on those sessions, we produced a proof of concept that addressed all of Tennet's challenges. Paulus: ‘What makes Tennet's puzzle complex are the many conditions that apply to each project. When do you start which project? The planners at Tennet need to be certain at the start that the necessary resources will be available. And those are primarily critical, technical roles that are in short supply. Besides, each project has a certain level of importance. One supports maintenance, the other the stability of the network. We therefore compiled a number of project criteria that are assessed by the work package planner when a schedule is prepared. This enables scenario-based thinking. Suppose you assign priority to new clients such as solar and wind farms and feed that in, the algorithm then rejigs the puzzle and produces a new planning schedule.’
optimal work package
The intelligent tool developed by SST and CQM, a web application, calculates the optimal work package using a mathematical model. This optimal package is based on the required and available resources. The tool enables users to specify particular KPIs when calculating the work package. That way, the impact of strategic choices can be measured immediately. The work package planner indicates by means of a score to what extent projects have been planned as optimally as possible. Tennet can therefore determine its optimal work package based on its strategic choices far more quickly and effectively.
‘The perspective of the planner is changing’, according to Van der Werff. ‘It can generate different schedules very quickly, depending on the variables you feed into it. You can then tinker with the results by calculating them. All of that could not be done manually, it is too time consuming.’
The key benefits of the new solution for Tennet are time savings and efficiency. ‘The smarter the way that the planning puzzle can be put together, the more projects Tennet can carry out. And the more efficiently Tennet can work, the more it can boost the energy transition.’ Another plus: greater transparency in the scheduling process. ‘Employees have their own way of doing manual planning, so you get different schedules. The new tool makes that a thing of the past.’
smart forward planning
Looking ahead gives you more control. The new solution means Tennet is better able to anticipate future scenarios. ‘The tool allows you to analyse which technicians you will need in the coming five years. If you already know that, you can start training them now. There is more time to analyse, to work smart, and to plan ahead.’
Part one of the tool is ready. Tennet's end users are learning how they can play their part as efficiently as possible. ‘It will take time to get used to the new way of working. Less manual work, more analysis. We have to be patient during this transition.’
Once they have embraced the new way of working, iterations that take the solution to the next level may follow. De Groot: ‘A product like this is never completely finished. It remains an interactive process.’ Even so, he already sees the smart work package planner as an asset for Tennet. ‘It expands our options. That is promising.’ It is the result of a constructive partnership with SST and CQM, he adds. ‘Effective communication, good harmonisation of expectations. I look back on the project with great satisfaction and look forward to the next chapter.’
The launch of our spin-off SST Labs!
Welcome ISO27001 certification
How pruning of neural networks results in reduction of energy consumption of Machine learning models.
Thanks to SST and CQM's solution, Tennet is able to put the pieces of its planning puzzle together in a smarter way