Embedded systems which support image processing have to cope with the challenging dilemma of the computing capacities versus the power consumption. A typical digital signal processing architecture, assigns an entire image to a single processor, but as real-time image processing applications have tremendous computational workloads and I/O throughput requirements, this leads to high frequency clocks and to an increase in electrical consumption. Next generation portable image products will benefit from the monolithic integration of photodetectors, analog-to-digital converters, digital processing, and data storage to improve the performance, efficiency, and cost. Advances in recent semiconductor technology have enabled the integration of CMOS image sensors with more digital processing circuitry thanks to amorphous silicon. Based on these new possibilities, the PACS project will propose a novel computing architecture able to support high level image processing algorithms near the sensor Focal Plane Array (FPA) with low-power consumption taking use of massive parallelism. The project will study an architecture made of an array of processors, each connected with a sub-array of pixels in the sensor plane. 3D integration of the detector plane stacked above the processing plane will be taken into account to define coarse grain processors integrating more complex functionalities and then offering much more flexibility. The regularity of SIMD structures and their hardware simplicity make them natural candidates to integrate a large number of processing elements with the FPA. Finally the PACS project has the ambitious goal to develop a new generation of smart sensors able to sustain much more complex image processing operations than existing ones keeping low power requirements thanks to a high level of parallelism. Smart pixel architectures offer important new opportunities for low cost, portable image processing systems. They provide greater I/O bandwidth and computing performance than systems based on classic microprocessors and are much more flexible than ASIC solutions. Effort to carry out the PACS project is estimated at 152 manxmonths and the total cost is settled at 1284 K€ (sub. 621 K€) for 30 months. This project targets the Theme 4 “Architectures innovantes” of the AF ANR directed towards fundamental researches.

The proposed project will deploy for the first time a new imaging cytometer platform capable of detecting minute quantity of micro-organisms in industrial and environmental waters. The platform is based on the integration of proprietary technologies available to the consortium partners: an automatic water concentration cartridge combined with a microfluidic cell will provide an adequate sample to a newly designed fluorescence image cytometer whose readings will be recorded and processed using a proper software interface. It will be validated for quantifying Legionella and Escherichia coli (E. coli) population within 120 minutes from obtaining the sample, overcoming in this way the main disadvantage of traditional methods used in laboratories, i.e. long time-to results which can currently last up to 12 days in the case of Legionella and 1 day for E. coli. The targeted detection limit will be 10-100 cells/L and 5-20 cells/100 mL for Legionella and E.coli, respectively. Also, the new imaging cytometer will have a portable form, a size similar to a smart-phone, which will increase its versatility and widen the possibilities of onsite applications. The relevance of the project is clear when one thinks about the high risk of legionellosis in some specific industrial environments, such as cooling waters, evaporative condensers and air conditioning systems and the fact that E. coli is one of the faecal pollution index commonly analyzed for monitoring the presence of waterborne pathogens and hence the quality of bathing waters. From a market perspective, more than 7 million of Legionella analyses are performed annually in Europe while E. coli level is included in all bathing water regulations in different EU countries. CYTO-WATER clearly falls into HORIZON 2020 topic WATER-1-2014/2015: Bridging the gap: from innovative water solutions to market replication and addresses Water Framework Directive (2000/60/EC) and in the Bathing Water Directive (2006/7/EC).