The Intelligent Systems Engineering Lab
The Department of Industrial Engineering and Management
Ben-Gurion University of the Negev
11/5/2023: Positions opened for graduate students in two projects 1) Learning advanced motion profiles using reinforcement learning 2) Advanced data analysis for assessing motion rehabilitation. Interested students are welcome to apply. Please send letter of interest, grade transcript and CV to: sigalbe [at] bgu.ac.il
The Intelligent systems engineering laboratory (ISEL) in the Department of Industrial Engineering and Management, at the Ben-Gurion University of the Negev, focuses on robots and intelligent systems capable of dexterous motion. We use machine learning and data-driven models for the synthesis of robotic motion and for the analysis of human motion.
Agriculture: We develop robotic systems with deep learning capabilities, advanced vision, and sensor fusion, which will enhance both production yield and sustainability. Upper-limb rehabilitation: We develop robotic and virtual reality-based systems for assisting motor recovery following stroke harnessing motor control theories. The digital factory: We develop methods for integrating robots in the digital factory.
We are looking for excellent undergraduate and graduate students interested in applying data-driven models and machine learning algorithms for developing robots and motion rehabilitation methods.
Lab director: Prof. Sigal Berman
Selected research projects
Error Augmentation for Upper Limb Rehabilitation for Stroke Survivors
Stroke is a leading cause of long-term sensorimotor disability with deficits in upper limb function persisting into the chronic stage in a large proportion of stoke survivors. This may partly be due to the limited effectiveness of current upper limb rehabilitation interventions. A stronger focus on training aligned with addressing the underlying causes of the motor control deficits rather than on only behavioral output, may be essential for significantly improving treatment outcomes. Multiple efforts have been initiated for augmenting rehabilitation following stroke with the advent of advanced technology, such as robotics or virtual reality (VR). One of the explored possibilities is based on error augmentation (EA). In EA treatment, subjects are provided with feedback that enhances their motor errors. This is usually done with distorted visual feedback and, using robots, by providing haptic feedback based on motion error. EA has resulted in positive effects for lower limb locomotor training but results for upper limb training are inconclusive. The current project investigates the effects of EA treatment for upper limb rehabilitation in subjects with stroke. We are developing training protocols based on motor control theory to re-mediate underlying movement deficits.
Funding: the Helmsley Charitable Trust through the Agricultural, Biological and Cognitive Robotics Center of Ben-Gurion University of the Negev and BGU Multi-disciplinary brain research.
Perception and action in remote and virtual environments
Telerobotic control and virtual reality (VR) environments share many similarities, e.g., the inherent spatial and temporal separation between perception and action. Moreover, VR is used in many telerobotic systems for locally closing the operator’s control loop facilitating separate operation of the remote and operator sites. Understanding the perception-action mechanisms and their applicability for telerobotic control and virtual environments is expected to facilitate the development of objective measures of system transparency, improved interface designs, along with improved operator training routines. Moreover, both remote and virtual environments, offer a unique platform for examining the mechanisms that underlie perception and action. In reach to grasp motion, a reaching movement is performed in free space, while the goal configuration is constrained by the configuration afforded by the object and functional task. Contact with the object is made at the end of the movement. Object characteristics, e.g., size, weight and texture, determine the finger aperture required for a successful grasp. Unlike grasping, the goal of reaching movements is in free space and the final hand orientation is typically not constrained. During reaching movements, when obstacles are present, movement trajectories are adjusted in order to avoid a potential contact with the obstacle. In the current project, we examine the influence of the interface, object and obstacle size, transmission delays, and viewing conditions, on motion in virtual and telerobotic environments.
Related publications:
O. Afgin, N. Sagi, I. Nisky, T. Ganel, S. Berman, 2017, Visuomotor resolution in telerobotic grasping with transmission delays, Frontiers in Robotics and AI-Bionics and Biomimetics, https://doi.org/10.3389/frobt.2017.00054 .
A. OzanaS, S. BermanC, and T. GanelPI, 2018. Grasping trajectories in a virtual environment adhere to Weber’s law. Experimental Brain Research, 236(6):1775–1787 .
A. Milstein, T. Ganel, S. Berman, I. Nisky, 2018. Transparency of grasping via a robot-assisted minimally invasive surgery system. IEEE Trans Human Machine Systems, 48(4): 349 - 358.
Industrial Internet of Things (IIOT) communication integrated in the telerobotic system
Funding: the Helmsley Charitable Trust through the Agricultural, Biological and Cognitive Robotics Center of Ben-Gurion University of the Negev.
Robotic Medjoul-date thinning
Fruitlet thinning is an important task of Mejdoul-date cultivation. Precise and scrutinized thinning is an essential stage toward attaining high quality yields. Currently, thinning is a labor-intensive process and automation of this process is essential to enable fruitlet thinning of larger plots with reduced man power. We are part of a multi-disciplinary team working together on the development of robot for Medjoul-date thinning. We are in charge of motion planning and fruitlet cluster identification. We are working with dynamic motion primitives, learning motion parameters using deep learning and kernel estimation methods. We are applying stochastic models for data generation. Using Bayesian methods for learning model parameters.
Funding: the Israeli Ministry of Agriculture & Rural Development and the Israeli plant council.
Robotic assembly planning for deformable objects
The BGU IEM ISEL team focus on assembly automation of products with deformable objects, developing heuristics for effective sequence planning of assembly activities and new methods for planning and cooperation between the robotic manufacturing cell components.
Related publications:
S. Ben-David, R. Shneor, S. Zuler, Z. Mann, A. Greenberg, S. Berman, 2021. Simulation-Based Two Stage Sequencing of Robotic Assembly Operations with Deformable Objects, 17th IFAC Symposium on Information Control Problems in Manufacturing (INCOM), Budapest (ZOOM), June 7-9.
R. Shneor, S. Berman, 2021. Robotic Manipulation: An Industry-Implementation Oriented Categorization, 26th International Conference on Production Research, 26th International Conference on Production Research, Taichung, Taiwan (ZOOM), July 18-21.
S. Ben-David, S. Berman, 2021. A multi-objective fitness function for sequencing robotic assembly operations with deformable objects using a genetic algorithm with constraint satisfaction, 26th International
Funding: the Israeli Innovation authority as part of the Robotics for Industry consortium.
AVR4Nano- Acoustic-Visual mobile Robotic manipulator for application of Nanostructures in agriculture
A major challenge in agricultural crop production, is containing and treating biotic and abiotic stresses in order to prevent epidemics, and to avoid yield losses. To significantly improve the capabilities of active compound application we are developing an innovative robotic solution for application of nanostructures in agriculture. Such a system offers a groundbreaking new direction in precision agriculture facilitating efficient application of nanostructures which encapsulate active compounds to their target sites on the plant. In a previous project we developed a robot-based system for disease identification in pepper plants. Automatic disease identification and eradication are expected to significantly improve yield, profitability, and sustainability.
Related publications:
N. Schor, A. Bechar, T. Ignat, A. Dombrovsky, Y. Elad and S. Berman, 2016, Robotic disease detection in greenhouses: combined detection of Powdery Mildew and Tomato Spotted Wilt Virus, IEEE Robotics and Automation Letters (RA-L) (and IEEE International Conference on Robotics and Automation - ICRA)).
N. Schor, S. Berman, A. Dombrovsky, Y. Elad, T. Ignat and A. Bechar, 2016. Development of a robotic detection system for greenhouse pepper plant diseases, Precision Agriculture, 18)3(:394–409.
Funding: the Israeli Ministry of Science and Technology (MOST).
ENHANCE - Enhancing brain plasticity for sensorimotor recovery in spastic hemiparesis
Spasticity is characterized by a velocity-dependent increase in tonic stretch reflexes in muscles at rest and during movement. In subjects with stroke, spatial threshold control of muscle reflexes is diminished. The reduced capability leads to abnormal muscle activation patterns i.e., excessive co-activation, as well as spasticity and weakness within well-defined spatial (angular) zones. The ENHANCE project is tri-national (Canada, India, Israel) effort harnessing advanced technologies such as virtual reality and Transcranial direct current stimulation (tDCS) for motor rehabilitation flowing stroke. Our team is in charge of analyzing the kinamtic data collected as part of the project. We are using advanced stochastic model for assessing the effects of spasticity on motion kinematics.
Related publications:
M. F. Levin, M. C. Baniña, S. Frenkel-Toledo, S. Berman, N. Soroker, J. M Solomon, D. G. Liebermann, 2018, Personalized upper limb training combined with anodal-tDCS for sensorimotor recovery in spastic hemiparesis: study protocol for a randomized controlled trial, Trials, 19:7, https://doi.org/10.1186/s13063-017-2377-6 .
I. Davitowitz, Y. Parmet, M. C. Baniña, S. Frenkel-Toledo, N. Soroker, J. M. Solomon, D. G. Liebermann, M. F. Levin, S. Berman, 2019. Effects of spasticity on upper limb movement in patients with stroke using stochastic spatiotemporal modeling, Neurorehabilitation and Neural Repair, 33(2): 141–152. [Link]
Funding: the Israel Science Foundation (ISF) , the Azrieli foundation, the Canadian Institutes of Health Research (CIHR) and the International Development Research Center (IDRC)
Inferring human intent in remote-control scenarios
The ability of inferring human intent can be instrumental in many human-robot interaction scenarios. Inferred intentions can be at both tactical and strategic levels. As remote operation becomes more prevalent, complex new interaction scenarios are emerging, e.g., a human remotely controlling a robot which interacts with another robot. The addition of a remotely controlled robot present further challenges to the already challenging task of inferring human intent. We have built in simulation and with hardware a robotic table soccer game where two robots facing each other hit a ball placed within a game court to form a test-bed for idenfitying human intent in remote scenarios. One robot is controlled autonomously using machine vision and learning algorithms, and the other is controlled by a remote human operator.
Funding: the Helmsley Charitable Trust through the Agricultural, Biological and Cognitive Robotics Initiative and the Paul Ivanier Center for Production Management at Ben-Gurion University of the Negev
Undergraduate
364-1-1011 Introduction to Industrial Engineering and Management
Fall 2020
Basis of IEM and Production revolutions.
364-1-3101Modeling and analysis of cyber-physical systems
Spring 2019
Modeling, design and analysis of cyber-physical systems.
Graduate
364-2-5421 Intelligent automation systems
Fall 2018
Foundations of intelligent system modeling and design.
Course project: robotic table soccer
2015 project - Autonomous transportation [project description]
2016 project - Plant disease detection (turn a leaf for detecting Powdery Mildew)
2019 project - Robot soccer (Unity simulation by Lily Sror and Ben Rachmut)
Principal investigator
Sigal Berman
Professor, Industrial Engineering and Management
Office: room 247, building 16, Marcus campus
Phone: 972-8-6479785
Email: sigalbe[at]bgu.ac.il
Research interests: robotics, human motor control, stroke rehabilitation
Students
Bar Shvarzman
MsC student, Industrial Engineering and Management
Office: room -126, building 16, Marcus campus
Email: barshv[at]post.bgu.ac.il
Project: Automatic planning of robotic wiring
Ran Shneor
PhD student, Industrial Engineering and Management
Office: room -126, building 16, Marcus campus
Email: shneorr[at]post.bgu.ac.il
Project: Automatic process planning for deformable objects
Alumni
Shai Chereshnia
MsC student, Industrial Engineering and Management
Office: room -126, building 16, Marcus campus
Email: shaiche[at]post.bgu.ac.il
Project: Extracting production details from product data
Zohar Karni
MsC student, Industrial Engineering and Management
Office: room -126, building 16, Marcus campus
Email: karniz[at]post.bgu.ac.il
Project: Learning motion for harsh terains
May Regev
MsC student, Industrial Engineering and Management
Office: room -126, building 16, Marcus campus
Email: mayre[at]post.bgu.ac.il
Project: Extracting Medjoul date features for robotic thinning
Edo Moran-Wexler
MsC thesis: Identifying grape features from robotic spraying, 2022
Currently: Data scientist, Explorium
Shahar Agami
MsC thesis: Error-augmentation for upper limb rehabilitation for stroke survivors, 2021
Currently: Data Analyst, DoubleVerify
Shir Ben-David
MsC thesis: Intelligent scheduling of robotic actions for deformable objects, 2021
Currently: Data scientist, Avantis (startup).
Omer Dadnon
MsC thesis: Learning robotic motion for Nano particle grape spraying, 2021
Tal Shoshan
MsC Thesis: Medjool-Date Fruit Bunch Segmentation and Parameter Estimation using Deep Learning, 2019
Lily Sror
MsC thesis: A virtual reality-based training system for error-augmented rehabilitation treatment in patients with stroke, 2019
Currently: bi analyst, MyHeritage
Or Bar-Shira
MsC thesis, Learning motion primitive parameters for a Medjool date thinning robot, 2019
Currently: PhD student in the Faculty of Mathematics and Computer Science, Weizmann institute of Science
Hadar Lackritz
MsC thesis: Using Hellinger’s distance for quantifying the effects of spasticity following stroke on voluntary motor control, 2019
Currently: Data scientist, WeissBeerger
Isgav Davidowitz
MSc Thesis: Relationship Between Spasticity and Upper-Limb Movement Disorders in Individuals With Subacute Stroke Using Stochastic Spatiotemporal Modeling, 2017
Currently: Business Analyst at Facebook, working with Health-Tech startups on data driven growth
Specializes: Health-tech, Statistics, Machine Learning, Marketing technology
Email: isgav1[at]gmail.com / isgav[at]fb.com
Rotem Duani
MSc Thesis: Bayesian Estimation of Spatial Configuration Distribution, 2017
Currently: Data Scientist at Moovit
Specializes: transportation, statistics and Machine learning
Email: rotemduani40[at]gmail.com
Omri Afgin
MSc Thesis: Perception in bilateral telerobotics, 2016
Currently: Data Scientist at Bank Hapoalim, Innovation Division.
Specializes in Machine Learning, Big Data Analytics and Statistics.
Email: omri17[at]gmail.com
Nir Sagi
MSc Thesis: Assessing transparency of a gesture-based interface for programming by demonstration, 2016
Currently: Product manager at the Innovation group, Siemens PLM.
Specializes in graphic solutions (mainly AR & web based tools) for industrial use-cases
Phone: +972-50-7482981
Email: nir.sagi[at]siemens.com \ nirsagi10[at]gmail.com
Noa Schor
MSc Thesis: A robotic system for diseases detection of greenhouse peppers, 2015
Currently: data scientist at Bank Leumi, Big Data division.
Specializes in in machine learning and big data applications as well as statistical analysis.
Roi Reshef
MSc Thesis: Planning Reach-to-Grasp Motions for a Robotic Arm using Rapid-exploring Random Trees, 2014
Currently: Leading a team of algorithm engineers and researchers that does planning and decision making for GM’s autonomous driving project.
Specializes: in robotics, planning, deep reinforcement learning, software engineering
Email: roireshef[at]hotmail.com
Danny Eizicovits
MSc Thesis: Examination of constrained motion on a hemispherical surface, Industrial Engineering, 2010
Phd Thesis: Graspability maps: representation, generation and integration in both gripper design and path planning, Industrial Engineering, 2014
Currently: CTO, The Decade Group – Young Engineers Ltd.
Specializes in robotics, embedded programming, industrialization and mass production of electronic toys.
Phone: 972-528-379-416
Email: danny[at]algobrix.com
Omri Mendels
MSc Thesis: Biometric identification by hand motion signature, 2011 (Co-advisor: Helman Stern)
Currently: Senior Data Scientist at Microsoft.
Specializes in machine learning and human computer interaction.
Tal Merdler, MSc., Objective measures in rehabilitation, Industrial Engineering, BGU, 2011
Eti Almog, MSc., Comparison of segmentation methods and hand tracking optimization, 2011
Kiril Smilansky, MSc., Classification of compound hand gestures, 2012 (Co-advisor: Helman Stern)
Shani Talmor, MSc., Interfaces for hand gesture browsing systems, 2012
Maayan Yaacobovich, MSc.,Examination of human grasp motion during selective harvesting, 2012
Merav Shmueli, MSc., Longest common subsequence classifiers for hand gesture recognition, 2012 (Co-advisor: Helman Stern)
Peer Shoval, MSc., Automatic guided vehicle dispatching, 2012
Anat Hershkovitz Cohen, MSc., Reach-to-grasp motion using dynamic movement primitives, 2014
Post-doctoral fellows
Dr. Darya Frolova, Gesture recognition, 2010 - 2011
Dr. Gali Naveh, Collaborative manufacturing network for competitive advantage, 2011 – 2013 (Co-host Lior Fink)
Dr. Orit Raphaeli, Collaborative manufacturing network for competitive advantage, 2012 – 2014 (Co-host Lior Fink)
Dr. Arun Kumar Singh, Human motion prediction and analysis for shared and transparent control, BGU, 2014 – 2016 (Co-host Ilana Nisky)
Contact
The Intelligent Systems Engineering Lab (ISEL)
Room -126, building 16
The department of Industrial Engineering and Management (IEM)
Ben-Gurion University of the Negev
email: sigalbe[at]bgu.ac.il
Tel: 972-8-6479785