Dye-sensitized solar cells under ambient light powering machine learning: towards autonomous smart sensors for the internet of things

Light option for machine learning

30 September 2020

Highly efficient solar cells optimized for harvesting ambient light in indoor environments could power a smart Internet of Things (IoT) network.

A team that included Technical University of Munich scientists has developed cells that capture energy from fluorescence lights with efficiencies of up to 34%. They achieved this by developing a dye-sensitized photovoltaic cell tailor made for harvesting the wavelengths of light available indoors and by suppressing energy losses from electron-back transfer in the cell.

The cells could capture enough energy to power autonomous IoT devices with artificial-intelligence capabilities. A wireless network of these IoT devices could use intermittently available ambient light to power an artificial neural network that demonstrated machine learning, a type of computation usually carried out on large servers.

Photovoltaic cells attuned to capturing ambient light could power a new generation of smart IoT devices, the researchers conclude.

Supported content

References

1. Chemical Science 11, 2895–2906 (2020). doi: 10.1039/c9sc06145b