The carefully developed data processing algorithms used by AQMesh are fully traceable and fixed by version number, and, unlike other systems, have been achieved with no use of machine learning or artificial intelligence.
Improvements in sensor technology and our understanding of how sensors respond in different conditions (from datasets of AQMesh pods co-located with reference stations around the world) have allowed us to make step changes in the accuracy of data. Each of these steps is released as a specific algorithm version, with a version number which is stamped onto each reading: total traceability.
The corrections which make up each algorithm use fixed equations to give total repeatability and this approach has earned the approval of senior air quality experts. This differs fundamentally from Artificial Intelligence (AI) which has its place but, in our opinion, not here. We have tried using AI and have seen how readily calculations can be led by dominant factors beneath the sensor signal, such as environmental conditions.
So what if a new algorithm is released? You have no obligation to adopt it if you are part way through a study, but every AQMesh user, no matter how old their pod, is welcome to take advantage of a new algorithm, at no charge.