The 16th meeting of the Working Group (WG) 41 of the European Standardization Committee took place at the end of April A couple of intensive days with online meetings due to the Covid-19 measures taken by most European countries at that stage. A nice session of statistics and mathematics took place during those days, with the Chebyshev's inequality and the Bland–Altman plot as stars of these sessions.
16 experts from Belgium, Italy, Germany, Netherlands, Spain and UK got together to bring ideas to the important issue of dealing with the mathematics involved in comparing the results of olfactometric samples with those of an IOMS. A calculation of a correlation coefficient has several limitations and that is why more advanced approaches were discussed during these two days of intensive work.
Unfortunately, as a Linux user unwilling to give my phone number to Microsoft ;-), I was among the minority that could not follow the written discussions, that were done on Microsoft Teams, so I missed that part. Luckily, the meeting took place in an open platform, and we could follow the oral discussions, which were also very interesting.
The discussion went on the best way to test quantification performance criteria of an IOMS. Several presentations were made and lots of fruitful discussions took place along these 2 days of morning meetings. The question on how many olfactometric analysis must be carried out to test IOMS performance on odour quantification was raised by the Task Group (TG) 1/2. This task group has had multiples meetings to discuss different approaches for odour quantification performed by an IOMS. Each procedure was discussed during this meeting.
The first step proposed was to test if an IOMS is not a random machine. Second test dealt with testing performance of the IOMS and for this task the Chebyshev's inequality was first proposed. The following day, a very nice presentation proposed the use of Bland–Altman plots to check performance of an IOMS. It is very difficult to explain in just a few lines, some of the discussions that took place on mathematics. Also, we had to take an intensive course on statistics to follow the conversation, but to be honest, it was fascinating.
If you want to participate in this process within this WG, why don't you join in? You will get access to the around 200 documents of this WG41. There is always a place here for any expert interested on this subject. Feel free to contact your national standardization body in the case of CEN members to get in touch with this group. Even if you are not European, you may get access to the documentation published in this group, provided that your country has an organization member of the ISO. Just contact your national representative body and ask them to link you to the CEN Technical Committee 264 Working Group 41.
If you find this article interesting, you might also be interested in these articles:
- First Meeting of the CEN group dealing with gas sensors and pattern recognition.
- Calibration of o-sensors? Is that possible? Some inputs from the last meeting in Berlin, Germany
- WG41 meeting on Instrumental Odour Monitoring Systems in Bilbao, Spain
- WG41 meeting in Barcelona
- WG41 last meeting of 2019 in Delft