Spatially Referenced Sensing Platform for In-Situ Monitoring in Photobioreactors
Year: 2026
Type: article
Abstract: Photobioreactors (PBRs) are spatially heterogeneous systems in which variables such as light intensity and biomass concentration vary throughout the reactor volume. Fixed probes provide only localized measurements and therefore cannot reconstruct these spatial variations. Here, we present a compact floating sensing platform equipped with a magnetic localization subsystem. Three external anchor coils are sequentially activated, and the resulting magnetic-field magnitudes are measured by an onboard magnetometer. An Artificial Neural Network (ANN) is used to map the magnetic-field triplet to the probe position. The ANN is trained on Biot–Savart data and then exposed to experimentally derived magnetic-field uncertainty. The spatial discretization of the training dataset is selected by comparing magnetic-signature separability with the effective measurement uncertainty. Simulations show millimeter-scale localization accuracy within the region of interest. Experimental validation at 312 known positions in a laboratory-scale mock-up confirms position reconstruction from real magnetic-field measurements, with accuracy depending on the local discriminability of the magnetic signatures. The proposed platform therefore provides the localization capability required to associate onboard sensor measurements with spatial coordinates inside a photobioreactor.
Source: Sensors
DOI: 10.3390/s26227216
Authors: Marinela Gjestila(Conceptualization, Data curation, Formal analysis, Software, Visualization, Writing – original draft), Valentina Diolaiti(Investigation, Methodology, Validation, Writing – review & editing), Leonardo Sollazzo(Formal analysis, Visualization, Writing – review & editing), Nazim Aslam(Formal analysis, Writing – review & editing), Diego Bernardoni(Formal analysis, Writing – review & editing), Fabio Rossi(Formal analysis, Writing – review & editing), Donato Vincenzi(Conceptualization, Funding acquisition, Supervision)
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7154715288
