Sensor data + machine learning = state-of-the-art maintenance
Fingrid needed to take a big technological leap to modernise the maintenance of their transmission system. Data needed to be collected from substations to gauge the devices’ status and need for maintenance in real time. Different data communication solutions were needed to transfer the measurements to a cloud environment. In addition, the data had to be unpacked and arranged so that it could be used to manage the maintenance of the whole transmission system.
The designing of the sensors for collecting the data is a close effort between the sensor suppliers and the Solita team. The data collected by over 400 sensors installed in a single substation are mined and modelled using machine learning. The result is a real-time dashboard running on the Microsoft Azure platform, which displays the status of the substations and their devices to end users, like Fingrid’s maintenance specialists.
“The building of IoT analytics might sound simple, but when you take a look under the hood, you realise how many different things you need to do and what variety of skills you need, and that it’s beyond the capabilities of most regular engineers.”
– Marcus Stenstrand, Digitalisation Manager, Fingrid
“The assessment of substation maintenance needs and the system development of predictive maintenance require enormous volumes of data to be processed and analysed. Our high degree of specialisation in data science and data system development made it possible to build the comprehensive solution Fingrid needed.”
– Kimmo Koskinen, Account Director, Solita.