Eyal Weiss

Dr. Eyal Weiss


Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 September 2015 in Sensors
Verified
Journal Article
Open access
PDF file
RARoger Alimi
Roger Alimi
EWEyal Weiss
Eyal Weiss
TRTsuriel Ram-Cohen
Tsuriel Ram-Cohen
NGNir Geron
Nir Geron
IYIdan Yogev
Idan Yogev
A dedicated Genetic Algorithm (GA) has been developed to localize the trajectory of ferromagnetic moving objects within a bounded perimeter. Localization of moving ferromagnetic objects is an important tool because it can be employed in situations when the object is obscured. This work is innovative for two main reasons: first, the GA has been tuned to provide an accurate and fast solution to the inverse magnetic field equations problem. Second, the algorithm has been successfully tested using real-life experimental data. Very accurate trajectory localization estimations were obtained over a wide range of scenarios.

3 September 2007 in Sensors
Verified
Journal Article
Open access
PDF file
EWEyal Weiss
Eyal Weiss
BGBoris Ginzburg
Boris Ginzburg
TCTsuriel Ram Cohen
Tsuriel Ram Cohen
HZHovav Zafrir
Hovav Zafrir
RARoger Alimi
Roger Alimi
JSJacob Sharvit
Jacob Sharvit
The purpose of this paper is to present a system developed for detection andaccurate mapping of ferro-metallic objects buried below the seabed in shallow waters. Thesystem comprises a precise magnetic gradiometer and navigation subsystem, both installedon a non-magnetic catamaran towed by a low-magnetic interfering boat. In addition wepresent the results of a marine survey of a near-shore area in the vicinity of Atlit, a townsituated on the Mediterranean coast of Israel, about 15 km south of Haifa. The primarypurpose of the survey was to search for a Harvard airplane that crashed into the sea in 1960.A magnetic map of the survey area (3.5 km2 on a 0.5 m grid) was created revealing theanomalies at sub-meter accuracy. For each investigated target location a correspondingferro-metallic item was dug out, one of which turned to be very similar to a part of thecrashed airplane. The accuracy of location was confirmed by matching the position of theactual dug artifacts with the magnetic map within a range of ± 1 m, in a water depth of 9 m.

19 April 2024 in Electronics
Verified
Journal Article
Open access
PDF file
EWEyal Weiss
Eyal Weiss
SCShir Caplan
Shir Caplan
KHKobi Horn
Kobi Horn
MSMoshe Sharabi
Moshe Sharabi
This paper introduces a pioneering method for real-time image processing in electronic component assembly, revolutionizing quality control in manufacturing. By promptly capturing images from pick-and-place machines during the interval between component pick-up and mounting, defects are identified and promptly addressed in line. This proactive approach ensures that defective components are rejected before mounting, effectively preventing issues from ever occurring, thus significantly enhancing efficiency and reliability. Leveraging rapid network protocols such as gRPC and orchestration via Kubernetes, in conjunction with C++ programming and TensorFlow, this approach achieves an impressive average turnaround time of less than 5 milli-seconds. Rigorously tested on 20 operational production machines, it not only ensures adherence to IPC-A-610 and IPC-STD-J-001 standards but also optimizes production efficiency and reliability.

7 July 2023 in IEEE Transactions on Components, Packaging and Manufacturing Technology
Verified
Journal Article
Closed access
EWEyal Weiss
Eyal Weiss
Corrosion on electronic component terminations during assembly can lead to the failure of electronic devices. The terminations of electronic components are susceptible to corrosion when exposed to moisture and other corrosive agents during and before the assembly process. This corrosion can cause physical damage to the terminations, resulting in poor electrical contact and possible failure of the electronic component. In this paper, we present a case study where an automotive production line utilized Cybord’s AI-based inspection system to detect and prevent contamination in the soldering terminations of electronic components. The system interfaced with the vision system of pick-and-place machines in real-time and collected bottom-side images of all components placed on printed circuit boards (PCBs). The AI algorithm, based on a 3 billion component database, detected evidence of corrosion, mold, and other contaminants on each component and allowed the removal of poor-quality components from production. The reel was disqualified and sent to a lab for SEM-EDX analysis, which confirmed the findings of the AI algorithm, that the issue was evidence of oxidation contamination. The results of this case study demonstrate the effectiveness of using AI-based inspection in detecting and preventing contamination in electronic assembly, boosting the overall quality and reliability of the final product.

4 November 2009 in Sensors
Verified
Journal Article
Open access
PDF file
RARoger Alimi
Roger Alimi
NGNir Geron
Nir Geron
EWEyal Weiss
Eyal Weiss
TRTsuriel Ram-Cohen
Tsuriel Ram-Cohen
A detection and tracking algorithm for ferromagnetic objects based on a two stage Levenberg Marquardt Algorithm (LMA) is presented. The procedure is applied to localization and magnetic moment estimation of ferromagnetic objects moving in the vicinity of an array of two to four 3-axis magnetometers arranged as a check point configuration. The algorithms first stage provides an estimation of the target trajectory and moment that are further refined using a second iteration where only the position vector is taken as unknown. The whole procedure is fast enough to provide satisfactory results within a few seconds after the target has been detected. Tests were conducted in Soreq NRC assessing various check point scenarios and targets. The results obtained from this experiment show good localization performance and good convivial with “noisy” environment. Small targets can be localized with good accuracy using either a vertical “doorway” two to four sensors configuration or ground level two to four sensors configuration. The calculated trajectory was not affected by nearby magnetic interference such as moving vehicles or a combat soldier inspecting the gateway.

14 June 2023 in IEEE Transactions on Components, Packaging and Manufacturing Technology
Verified
Journal Article
Closed access
EWEyal Weiss
Eyal Weiss
Corrosion is a prevalent failure mode in electronic products. The initiation of failure often stems from pre-existing corrosion contamination on soldering terminations prior to assembly. This corrosion is further accelerated by environmental factors such as humidity, temperature, and acidity, ultimately leading to degradation of the board and failure during both post-assembly testing and the product’s lifespan. This study presents a method for the real-time, early detection of corrosion contamination on electronic components during the mounting process using pick-and-place technology. The method utilizes the correlation between light reflectance from soldering terminations during placement photography and the degree of corrosion present. Corroded terminations possess a rougher surface and pitting spots which result in different light reflectance compared to pristine terminations. This difference can be detected through AI forensic analysis of component images. The study presents an AI model that correlates termination finish with corrosion content and progression, and evaluates its performance on large-scale data. This study also presents a real-world case where corroded components were identified during the pick-and-place process, but later failed during in-circuit testing (ICT). The post-failure analysis, using scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDS) and cross-section analysis, confirms the accuracy of the AI failure predictions on multiple components with corrosion, during large-scale production. The proposed method has been implemented in multiple production lines, where it inspects all components without compromising throughput, and identifies contaminated components that are unsafe. The method has been tested on over 3.5 billion components, and has achieved an accuracy rate of over 99.5% in its predictions.