Geostats 2021 and more

So, another week’s gone. The last couple of weeks have been pretty hectic. On the 14th of June a little new member of the family arrived. Eskild is his name. Sweet and happy, but not sleepy at all. I guess he takes that on from his older brother.

Anyway, what this means is that I have been on parental leave since then. Thus, my working activity has been stagnant and will remain so until the start of August. This week however was (unfortunately since I couldn’t participate) the postponed week for geostats 2020. You know the pandemic 😔

I was scheduled for two presentations, which were prerecorded and shown at the conference. Thanks to some very much appreciated help from my co-authors Ingelise Møller and Thomas Mejer Hansen who were also part of the conference a discussion session for each presentation was also possible. I will put both presentations on my webpage shortly. If you also want to see the q and a session for each presentation I can firmly recommend registering for the conference and gaining access to all presentations and discussion sessions. I know at least what I will be up to when the two boys go to sleep in the next weeks.

I sincerely hope that it is possible to go ahead with Geostats 2024 in normal fashion. Such an interesting crowd of people and topics. Geostats 2016 was hands down the best conference I have ever attended.


New research added

Just wanted to let you know that I have added the presentation and abstract from the dwf2020 meeting on the webpage. Furthermore, I found an accepted abstract lying on my computer from last year. It was presented last year (2019) at the ‘sewer processes and networks’ in Aalborg. I mainly wrote the two page abstract as a conclusion on a small project I participated in with some good people at the consultancy company Niras. The project revolved around sewage pipe failures. We wanted to see whether we could establish a classification relationship between broken pipes and different pipe features such as age, material etc. Since I was only working as a subcontractor, my name could not put as the main author (company policy). Instead, Mads Paulsen ended up as the first author and also presented the findings at the conference. Made perfect sense since he handled the majority of the modelling. Had quite a bit of fun doing this project actually, and it presented a quite good case for applying machine learning. This is definitely not always the case even though it is being pushed heavily these days. It turned out that we indeed (maybe unsurprisingly) where able to predict some correlation between broken pipes and our proposes features. The abstract link is in the research archive as well.