One thing that seriously concerned me with these SHT31 evaluation boards was the long term shift in accuracy on the humidity side. I only test a particular type of sensor in minimum groups of three. When a sensor company is wanting to bid on a government contract it's usually pretty easy for me to secure at least three for testing. Although that's not the case with these evaluation boards, it was more for my own curiosity. The issue is approximately 3 weeks after receiving all three of these evaluation boards, I ran a test at 50% RH. All three were reading low by -1.2%, -1.7% and -1.9%. Almost a year later they were reading +0.5% to +0.8%. I'm at a loss to explain why the large shift over time. I'm not quite ready to call it "drift" because that's an insane amount of drift. I was so concerned, I sent my NIST reference instrument back to Washington D.C. to get it certified again after it was just certified 6 months before. It came back having tested just fine. In the hundreds of sensors I've tested over 20+ years, I never seen this much shift before over the course of a year. I was thinking of getting a fourth one and seeing how it compares right out of the box against these three units that have been running for almost a year. Some sensors do require a conditioning period when first used or what I refer to as exercising the sensor. I know that with some sensor manufacturers when they send you a sample for evaluation, sometimes these are not their best spec'ed sensors. Often they are sensors that are usable but don't quite pass QC 100% for some reason. I wonder if that's what these evaluation boards are. I never did ask Sensirion. The only thing I can think of is since these SHT31 sensors don't have a filter over them, they have become contaminated with dust over time.