Page 67 - tech fest 2025 ב״ש חוברת תקצירים

Book of Abstracts | 2025 Advanced Portable Cooler for Medical Blood Transfusions in IDF Combat Zones
EEE-A-11
Ofek David Pinzaru; ofekdavidp97@gmail.com Sagi Alterman; sagial@ac.sce.ac.il
Advisors: Dr. Efi Zemach1, Ms. Nataliya Dvoskin1 1SCE - Shamoon College of Engineering, Be’er-Sheva
During combat, the primary cause of preventable mortality among wounded combatants is hemorrhage. Blood transfusion as hemorrhage treatment during combat faces difficulties with transport and proper storage of the packed red blood cells (pRBC) to preserve their vitality. This project solved these difficulties by designing a portable active cooler, powered by the IDF’s radio transceiver Li-Ion rechargeable battery pack MR-2791. Portability in cooling was achieved by using two thermoelectric-cooler modules, which directly convert electrical power to heat-transfer extraction. Storage temperature was regulated by a control system incorporating an Arduino Uno microprocessor, which monitors the interior temperature and the battery power in real time, ensuring constant pRBC storage temperature while tracking the battery energy status.
Keywords: control system, heat transfer, packed red blood cells storage, thermoelectric cooling
Geoengineering Characterization of Quarry Materials Using Field-Based Infrared Hyperspectral Imaging and Neural Networks
EEE-A-12
Shirel Zigdon; shirelzig1121@gmail.com Daniel Fatahov; fatahovdanie12@gmail.com
Advisors: Dr. Isaac August1, Dr. Dagan Bakun Mazor1 1SCE - Shamoon College of Engineering, Be’er-Sheva
This project introduced a practical method for evaluating the mechanical properties of quarried rock fragments without physical examination or destructive lab tests. Instead of extracting rock samples and testing them in the lab, we used an advanced near-hyperspectral imager in the near-infrared (NIR) range to image rock fragments under natural sunlight conditions. The images contained rich spectral data, which we preprocessed, calibrated, and filtered to remove unreliable wavelength regions affected by atmospheric absorption. We then trained a neural network to predict important physical properties from the spectral data, such as density, water absorption, and uniaxial compressive strength. This field-based approach offers a fast, low-cost, and nondestructive alternative for remotely characterizing rock used in the construction industry.
Keywords: hyperspectral imaging, near-infrared, neural network, rock characterization, radiometric calibration
67