The local wastewater company VEAS wanted master students to apply their modelling skills to the urban drainage network in the Oslo area, and invited artificial intelligence, civil-, mechanical-, electrical- and chemical engineering master students to the VEAS Hackathon.
The VEAS plant is Norway's largest treatment plant, and a crucial contributor to the work of keeping the Oslo Fjord clean.
The VEAS tunnel transports wastewater from more than 600,000 inhabitants in Oslo, Bærum and Asker to the treatment plant at Bjerkås in Asker.
The digital Hackathon started 15 October 2020, and after 24 hours, the student groups, submitted their answer and presented their solutions in Zoom.
All of the students are from the Master’s Programme Degree in Applied Computer Science and Information Technology (ACIT) at OsloMet.
The winner team
A panel of experts from Oslo, Bærum and Asker municipalities and VEAS evaluated the results, and the winner team was a group with the students Marit Øye Gjersdal, Andreas Kjernlie, Sergio Alejandro Sotres Romero and Pål Anders Owren.
Marit specializes in Applied Artificial intelligence, Andreas in Robotics and Control, Sergio in Mathematical modelling and Scientific Computing and Pål in Data science at the Master’s programme.
In their solution, the students identified the drainage zones that have the largest excess of water into the tunnel during rainy periods.
‘We can calculate linear correlation measures between the amount of rain and the inlet ﬂow for diﬀerent geographical zones in the three municipalities,’ the students suggested.
Then they continued with the analysis:
‘First, we analysed the raining data for Oslo, Bærum and Asker.’
‘We considered a visualization per day. We repeated the same procedure for the inlet ﬂow data ﬁles considering a one-day time-step (24 hours).’
‘We calculated the correlations between the diﬀerent raining regions and the inlet ﬂow regions, and catalogued the types of correlations between the diﬀerent zones.’
‘From here, we could determinate if the Inlet ﬂow increases due to the rain water.’
The students suggested further:
- Using more time to analyse the data ﬁles to obtain better correlations from a large number of regions.
- Appling the correlation between rainfall and inlet ﬂows to identify conﬂicted areas.
- Measuring incoming ﬂow in conﬂicted areas and use of AI (Artificial Intelligence) connected to the sensors to predict overﬂows, open and close inlets and regulate use of chemicals.
- Using inlet pipes with low water traﬃc as possible water reservoirs during heavy loads, to prevent overﬂow.
The winner group got an award of 10 000kr from VEAS
Associate professor Tiina Komulainen from OsloMet and postdoc Jakub Roemer for VEAS organized the Hackathon in cooperation with Oslo, Bærum and Asker municipality, and University of South-Eastern Norway and University of Agder.
The winner team on the picture above: From left Sergio Alejandro Sotres Romero, Marit Øye Gjersdal, Pål Anders Owren and Andreas Kjernlie. Photo by Karin Doan.