Review on Portable Airthings Corentium Home Radon Detector 223: Lightweight, Easy-to-Use, USA Version, pCi/L by Shane Bullion

I'm a researcher at the National Laboratory. I bought a Corentium home radon detector (Airthings 223) in September 2018 and my friend bought the same one in October 2018. We compared the results of these devices in November 2018 for 3 weeks in the basement at the same time. Both devices were closed to each other: a gap of 2 cm, as shown in the figure. It was assumed that the readings would be the same within a reasonable tolerance, however the two devices showed wildly different readings. One was 1.72 pCi/L and the other was 3.18 pCi/L for the long term average. This means that the accuracy of the device is very low or the calibration was completely wrong. I am so disappointed in Airthings and do NOT recommend this radon detector. After contacting the manufacturer I will update their answer. Update 03/09/2019: Customer service was good. They react immediately and replace the product with a new one. I wanted to recalibrate an item to compare the value before and after the calibration. However, instead of recalibrating, they could only do a swap. I'm currently testing a new article and will update the result soon. Updated 11/14/2020. Sorry for the late update. This replaced new airthing was almost the same value as the item on the left of the image uploaded, in the same location and condition. Radon was recently measured twice using new Airthings 223 and Model 1028 replacements from Sun Nuclear Co. at the same time every 48 hours in my basement. When first measured, the Airthings 223 had 0.24 pCi/L and the 1028 model 0.5 pCi/L under a frequently opened window. In the second measurement with the windows closed, Airthings had 0.74 pCi/L and Model 1028 had 1.6 pCi/L. The Model 1028 is calibrated annually by Sun Nuclear Corporation of Florida. Then it looks like the real cost is about twice the cost of my new Airthings 223. The manufacturer's calibration is totally wrong and they are not for recalibrating.