A Monitoring Framework with Integrated Sensing Technologies for Enhanced Food Safety and Traceability.

A novel and low-cost framework for food traceability, composed by commercial and proprietary sensing devices, for the remote monitoring of air, water, soil parameters and herbicide contamination during the farming process, has been developed and verified in real crop environments. It offers an integrated approach to food traceability with embedded systems supervision, approaching the problem to testify the quality of the food product. Moreover, it fills the gap of missing low-cost systems for monitoring cropping environments and pesticides contamination, satisfying the wide interest of regulatory agencies and final customers for a sustainable farming. The novelty of the proposed monitoring framework lies in the realization and the adoption of a fully automated prototype for in situ glyphosate detection. This device consists of a custom-made and automated fluidic system which, leveraging on the Molecularly Imprinted Polymer (MIP) sensing technology, permits to detect unwanted glyphosate contamination. The custom electronic mainboard, called ElectroSense, exhibits both the potentiostatic read-out of the sensor and the fluidic control to accomplish continuous unattended measurements. The complementary monitored parameters from commercial sensing devices are: temperature, relative humidity, atmospheric pressure, volumetric water content, electrical conductivity of the soil, pH of the irrigation water, total Volatile Organic Compounds (VOCs) and equivalent CO2. The framework has been validated during the olive farming activity in an Italian company, proving its efficacy for food traceability. Finally, the system has been adopted in a different crop field where pesticides treatments are practiced. This has been done in order to prove its capability to perform first level detection of pesticide treatments. Good correlation results between chemical sensors signals and pesticides treatments are highlighted.

Towards Enhancing Traffic Sign Recognition through Sliding Windows.

Automatic Traffic Sign Detection and Recognition (TSDR) provides drivers with critical information on traffic signs, and it constitutes an enabling condition for autonomous driving. Misclassifying even a single sign may constitute a severe hazard, which negatively impacts the environment, infrastructures, and human lives. Therefore, a reliable TSDR mechanism is essential to attain a safe circulation of road vehicles. Traffic Sign Recognition (TSR) techniques that use Machine Learning (ML) algorithms have been proposed, but no agreement on a preferred ML algorithm nor perfect classification capabilities were always achieved by any existing solutions. Consequently, our study employs ML-based classifiers to build a TSR system that analyzes a sliding window of frames sampled by sensors on a vehicle. Such TSR processes the most recent frame and past frames sampled by sensors through (i) Long Short-Term Memory (LSTM) networks and (ii) Stacking Meta-Learners, which allow for efficiently combining base-learning classification episodes into a unified and improved meta-level classification. Experimental results by using publicly available datasets show that Stacking Meta-Learners dramatically reduce misclassifications of signs and achieved perfect classification on all three considered datasets. This shows the potential of our novel approach based on sliding windows to be used as an efficient solution for TSR.