wisedroid wrote:
That’s brilliant - thank you so much for looking into this. Will be interesting to hear the result. Happy to do some testing if needed
Hey @wisedroid,
I have an answer and now a better understanding of how our sensor algorithm work for myself. Thanks for prompting this question on the community :).
See my reply below:
The sensors onboard your machine have a programmable algorithm that compensates for a number of gaseous particles. Some of these particles can provide a false positive to other readings. In your case, a spike in the Humidity has reflected as NO2. As this embedded compensation is set at a steady rate, it wouldn’t usually be triggered in standard scenarios. Meaning, a lower spike in Humidity would not of triggered the machines reading and prompt a reaction. Unlike what you have experienced. This compensation to the algorithm is based off of a number of factors, including historic connection data and laboratory tests.
After reviewing the data your machine has sent to the cloud, the team were able to identify that the Humidity change is such that is breaches this compensation level, and therefore triggers a reaction from your machine, which is in turn is being recorded as a different gaseous source. NO2. As the machine is in the AUTO function, it reacts to this detection and increased the airflow.
As the compensation level in these particles is programmed at build, it is not something that can be amended for individual customer circumstances. This means that in your case, when you open the door, as the increase in the Humidity from the outside environment is such, if it remains at the level it is, it will always trigger this reaction. As the outside humidity changes over time and throughout the year this isn't something I’d expect you to see all the time however.
There are a couple of things I can suggest for you to try, such as: Changing the machines location in the home, away from the airflow of the door and monitoring the outside humidity compared to that of the inside. Knowing though trial and error the conditions on when this reaction will trigger could help inform the machines reaction in the future.
As we continuously look to improve our technology and their effects on the environment around them, this occurrence has been noted by the team and will be looked at for future developments, in laboratory testing and further market research. Thank you for highlighting it as a use case. For now, should you wish to share any of the data you gather on when this triggers, please feel free to send it to me on a PM and I will feed it back to our teams.
As this question has no doubt benefited the community by prompting a wider understanding on my side and allowing a medium to explain our algorithms, I’ve award you the Community Builder badge. This is given to someone that implement change across the community.
I generally hope this helps answer your question and has given a small insight into how our algorithms work.
Many thanks,
Stuart
