General information
Offer title : PhD (M/F): Real-time water quality monitoring system (H/F)
Reference : UMR5129-MARCLO-103
Number of position : 1
Workplace : GRENOBLE
Date of publication : 26 September 2025
Type of Contract : FTC PhD student / Offer for thesis
Contract Period : 36 months
Start date of the thesis : 1 January 2026
Proportion of work : Full Time
Remuneration : 2200 € gross monthly
Section(s) CN : 08 – Micro and nanotechnologies, micro and nanosystems, photonics, electronics, electromagnetism, electrical energy
Description of the thesis topic
Real-time water quality monitoring system.
– Implementation of a multiplexed data acquisition and processing system and creation of a database for the various pollution sensors with a view to training online (non-embedded) models in the first instance.
– Development of a machine learning algorithm based on the study database and fed with data available in the literature. The model will first be developed online and then reduced and optimized for integration into embedded electronics, taking into account performance constraints as well as energy autonomy.
– Evaluation of system performance on samples under real conditions.
Work Context
The proposed work involves developing a real-time water quality monitoring system. Potential applications will initially focus on drinking water distribution networks. The main sources of water pollution are relatively well documented in the literature. Approximately 50% of water pollution is linked to the agricultural sector. The objective is to be able to monitor in real time the concentrations of ammonium ions (NH4+) and nitrates (NO3-), which represent the majority of polluting ionic species, as well as toxic heavy metal ions (Pb2+). The multiplexed detection system proposed for the study is based on the cointegration of several sensors that are selectively sensitive to the targeted pollutants, as well as temperature and pH measurements to enable calibration of the measurements. The development of specific instrumentation for the acquisition and processing of data in situ, as close as possible to the point of capture, is also a very important objective for reliable and self-calibrated water quality measurements.
Scientific and technological barriers:
Sensor selectivity is the first important barrier to overcome in order to perform quantitative analyses for each pollutant and avoid ionic interference between the different sensors used in the project. Sensor selectivity is already being studied in the laboratory using approaches based on the functionalization of sensitive membranes.
Original contributions expected:
Software approaches involving the development of machine learning algorithms are of real interest in improving the accuracy of water quality measurements, particularly in identifying, accounting for, and neutralizing ionic interference. The second key element of embedded intelligence lies in the sensor’s ability to self-calibrate and, in particular, to adapt its responses and models according to sensor aging and the (sometimes significant) variability of external conditions.
Recent work has demonstrated the strong potential of combining Internet of Things (IoT) devices with artificial intelligence in the cloud, but to date no solution has proposed embedding artificial intelligence in the sensor. The main challenges of the proposed mission are the long-term performance of the system designed on an IoT basis coupled with artificial intelligence, which is initially remote and then integrated as close as possible to the sensors in a second phase.
Constraints and risks
None
