Environmental monitoring (water and air) using PiSENS, a low-cost, portable, Raspberry-Pi based colourimeter

Here we are presenting a portable Raspberry-Pi based colourimeter. The optical signal provided, colour-change in this case, can be picked up by the camera of the portable device (PiSESNS) and processed automatically. To do so, we have written a fit-for-purpose code using the Python programming language. In addition, due to the versatility of the device, we have built two modes of PiSENS using materials from a DIY shop and pieces created with a 3D printer. The first one, called PiSENS-A, was set up for continuous measurement of NO2 in air. This
system is based on the Saltzman reaction by which the absorbance solution changes colour because of its interaction with the gas. The device allows data acquisition in a continuous way, this is with high temporal resolution. Due to its portability and low-cost construction, a high spatial resolution can be achieved as well. Successful data collection has been got for periods of 24 hours and 48 hours. The R2 that we got for 7 points calibration was 0.944, with a LOD of 7x10-5 ppm. The precision as relative standard deviation percentage was 2.8% (RSD%), and 6.4% was the standard error percentage (SE%).
The second one, PiSENS-O, was set up for one-shot measurements for the determination of specific analytes in water. In this case, ionophore-based polymer membranes with optical signals, optodes, are used for the analysis. As proof of concept, we successfully measured sodium (Na+) and nitrate (NO3-) in water. The results show a good adjustment between the theoretical sigmoidal curve and the calibration points. For both ions, the average accuracy measured as SE% was 7.4%, and the precision as RSD% was between 2-12.5% and 5-27% for Na+ and NO3- respectively.
The simplicity of the device in addition to its low construction and maintenance cost provide PiSENS with a lot of potential for its application in rural and remote locations. Particularly, where expensive equipment is not affordable or cost-effective. Furthermore, the data communication capabilities of Rasberry Pi such as WiFi and Bluetooth would allow the integration of multiple devices to create large networks that could assist with live decision-making or the creation of high precision models.