Natural illumination has an important place in home automation applications. Among
other advantages, it contributes to better visual health, energy savings, and lower CO2 emissions.
Therefore, it is important to measure illuminance in the most accurate and cost-effective way. This
work compares several low-cost commercial sensors (VEML 7700, TSL2591, and OPT3001) with a
professional one (ML-020S-O), all of them installed outdoors. In addition, a platform based on the
Internet of Things technology was designed and deployed as a centralized point of data collection and
processing. Summer months have been chosen for the comparison. This is the most adverse situation
for low-cost sensors since they are designed for indoor use, and their operating range is lower than
the maximum reached by sunlight. The solar illuminance was recorded every minute. As expected,
the obtained bias depends on the solar height. This can reach 60% in the worst circumstances,
although most of the time, its value stays below 40%. The positive side lies in the good precision of
the recordings. This systematic deviation makes it susceptible to mathematical correction. Therefore,
the incorporation of more sensors and data that can help the global improvement of the precision
and accuracy of this low-cost system is left as a future line of improvement.
