ISL | Impact through Systems Learning
About
The core focus of the Intelligent Systems Lab (ISL) is on developing novel reasoning and learning theories and algorithms, studying their impact on the wider community, and ensuring they are used in a responsible and safe manner. A primary aim of the ISL is to strengthen academia-industry linkages, moving cutting-edge technologies into production.
Staff
The ISL comprises a core group of academic and industry experts. Partnerships with for-profit institutions are established through the signing of Memoranda of Understanding (MoUs). Currently, the ISL has academic partnerships with researchers from Rutgers University in New Jersey, USA, and Kyungpook National University in Daegu, South Korea.
Application Spaces
Some of the work in the field of ISL is being applied to various areas, including policy and governance, robotics, SCADA systems, attack-resilient systems, digital twins, and AI-enabled software, such as vector databases, AI-enabled servers, and industrial operating systems (for more information, please refer to our GitHub repository)
Information Dissemination
The ISL disseminates information through scientific publications, conferences, seminars and media.
Ongoing Projects Catalogue
The following listings comprise projects that are not bound by an NDA (Non-Disclosure Agreement). Each project is licensed separately and assigned a Technology Readiness Level (TRL) ranging from 1 to 9. See more information about TRL here.
Attack Resilency Framework
Maintainer: Amir Mohammed
TRL: 6
This project is a software and technology based framework for the study of attack resiliency in control systems. Attacks defined here, spans DDOS, FDI and Replay.
Read more...
SimuNEX
Maintainer: Lee Bissessar
TRL: 7
SimuNEX is an AI-enabled high-fidelity simulation platform consisting of a physics engine, advanced control systems solver, HMI developer and test harness. It is adaptable for many applications.
Read more...
Quantitative Risk Framework for AI systems
Contributor: Samantha Deonarine
TRL: 6
Safety is necessary for LLMs because they have the potential to generate harmful or inappropriate content. This project aims to develop a quantitative risk reduction framework for linguistic reasoning machines.
Game Theoretic Approaches for Linguistic Reasoning Machines
Contributor: Derron Phagoo
Contributor: Giovanni Alexander
TRL: 6
This project aims to enhance machine reasoning capabilities, reducing hallucinations and disinformation with the use of a Multi-Agent structure studied in a Game Theoretic framework.
Industrial AI for SCADA Systems
Contributor: Dana Rattansingh
TRL: 8
This project is aimed at enabling SCADA systems with feature detectors, language models and retrieval augmented generation technologies for better decision support during fault, cyber-threat and abnormal situations.
Byte-Level Generative Networks: At Scale Universal Approximators are Number System Invariant
Maintainer: Craig Ramlal
TRL: 5
This project explores novel attention head mechanisms that result in subquadratic costs both at training and inference and 1/2 bit parameter sizes. We hypothesize that at scale, the performance of pre-trained, fune-tuned quantized models have the same perplexity of their floating point counterparts.
Federated Database Systems for AI Memory Management
Maintainer: Kevon Andrews
TRL: 2
This project aims to develop a universal fault-tolerant data orchestration system that is interoperable among various data-sources and has a low level of reasoning.
Research Grants
The following lists the research grants awarded to the ISL
01
Hospitech: Supporting in-hospital COVID-19 response through technology (Research and Development Initiative $120,000USD). Click here for further information
Status: Completed
02
CRP.3NOV23.06: Toward Faithful AI Systems: A Multi-Agent Unfalsified Framework ($11,500USD)
Status: Ongoing
Publications
Phillip, Micah, Arvind Singh, and Craig J. Ramlal. "Narrow Band Frequency Response Analysis of Power Transformers with Deep Learning." Energies 16, no. 17 (2023): 6347.
Mungal, M. J., A. Singh, C. J. Ramlal, and J. Colthrust. "Sensitivity analysis of the unit commitment problem to guide data acquisition investments in a small island developing state: A case study." Results in Engineering 18 (2023): 101191.
Hunte, Kyle, Craig J. Ramlal, and Jingang Yi. "Dynamic Path Planning for Multiple Robots Transporting Objects in a Deformable Sheet." IFAC-PapersOnLine 56, no. 3 (2023): 505-510.
Mohammed, Amir, Fasil Muddeen, Lincoln Marine, and Craig J. Ramlal. "The Exigency for Resilient and Cyber-Secure Critical Infrastructure in the Caribbean." West Indian Journal of Engineering 45, no. 2 (2023).
Manninen, Henri, Craig J. Ramlal, Arvind Singh, Jako Kilter, and Mart Landsberg. "Multi-stage deep learning networks for automated assessment of electricity transmission infrastructure using fly-by images." Electric Power Systems Research 209 (2022): 107948.
Mohammed, Amir, Fasil Muddeen, Craig J. Ramlal, and Lincoln Marine. "A Comprehensive Review of Fault Tolerant and Resilient Cyber-Secure strategies for Critical Infrastructure Protection." The Industrial Engineering and Management Journal 1, no. 1 (2022): 66-76.
Manninen, Henri, Craig J. Ramlal, Arvind Singh, Sean Rocke, Jako Kilter, and Mart Landsberg. "Toward automatic condition assessment of high-voltage transmission infrastructure using deep learning techniques." International Journal of Electrical Power & Energy Systems 128 (2021): 106726.
Ramlal, Craig J., Arvind Singh, Sean Rocke, Henri Manninen, Jako Kilter, and Mart Landsberg. "Toward automated utility pole condition monitoring: A deep learning approach." In 2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe), pp. 255-259. IEEE, 2020.
Ramlal, Craig J., Arvind Singh, and Sean Rocke. "Repetitive learning frequency control for energy intensive corporate microgrids subject to Cyclic Batch Loads." In 2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe), pp. 349-353. IEEE, 2020.
Mohammed, Amir, Craig J. Ramlal, Arvind Singh, Sean Rocke, and Daniel Goitia. "A simulation framework for controlled critical infrastructures subject to natural disasters." (2020).
Singh, Arvind, Sean Rocke, Akash Pooransingh, and Craig J. Ramlal. "Improving student engagement in teaching electric machines through blended learning." IEEE Transactions on Education 62, no. 4 (2019): 297-304.
Ramlal, Craig J., Arvind Singh, Sean Rocke, and Michael Sutherland. "Decentralized Fuzzy $ H_\infty $-Iterative Learning LFC With Time-Varying Communication Delays and Parametric Uncertainties." IEEE Transactions on power systems 34, no. 6 (2019): 4718-4727.
Dookie, Isa, Sean Rocke, Arvind Singh, and Craig J. Ramlal. "Evaluating wind speed probability distribution models with a novel goodness of fit metric: a Trinidad and Tobago case study." International journal of energy and environmental engineering 9 (2018): 323-339.
Current Researchers
Dr Craig Ramlal
Principal Investigator
Dr. Craig Ramlal earned his BSc(e), MASc, and PhD in Electrical and Computer Engineering split-site from KFUPM and UWI. He conducted impactful work such as coordinating open data strategies for Caribbean nations, leading projects with Ministry of Health and University of Florida for COVID-19 mitigation, and developing Industrial Diagnostic tools for Estonia's grid using deep learning techniques. His expertise lies in Intelligent Control Strategies, Game Theory and Artificial Intelligence.
Read more...
Dr Arvind Singh
Co-Investigator
Arvind Singh gained his B.Sc. in Electrical and Computer Engineering at the University of the West Indies in 2003. Subsequently he went on to study at the University of British Columbia where he obtained his Master’s and Doctoral degrees in 2006 and 2009 respectively. His research focuses on Condition Monitoring of Power Transformers and Electrical Machines, Renewable Energy, Large Scale Energy Management Systems and the Smart Grid. He is also involved in the design complex system simulation for modeling infrastructure interdependencies and emergency responses in times of disaster. In addition to these he is interested in developing pedagogical tools for music and language.
Kevon Andrews
Research Engineer
Kevon Andrews, with 15+ years in software engineering, is an Electrical and Computer Engineer from The University of West Indies, St. Augustine, holding a B.Sc. (2003) and M.Sc. (2006) in Communication Systems. He's delivered software solutions across industries like fisheries, agriculture, and medical. Since 2008, he's been an Engineer at the same university, co-leading projects such as FEWER, CC4FISH, and mFisheries. Currently, he focuses on software systems research, especially in AI.
Ravi Deonarine
Research Engineer
Ravi Deonarine is an Engineer II at the Department of Electrical and Computer Engineering. In addition to research and development projects he is responsible for the department’s online systems which includes several web applications that support the department operations. He also supervises undergraduate final year projects and has previously lectured in Telecommunications in both the undergraduate and postgraduate Electrical and Computer Engineering programs.
Amir Mohammed
Researcher
Received his BSc and MASc in Electrical and Computer Engineering with a major in Control Systems from the University of the West Indies (UWI), St Augustine. He is currently a PhD Candidate in Electrical and Computer Engineering at the UWI. His interests include Industrial Cyber Physical Systems, Resilient Control, Fault Tolerant Control and Artificial Intelligence.
Lincoln Marine
Researcher
He earned a BSc in Electrical and Computer Engineering focused on Control and Communication Systems from UWI, St. Augustine, with a keen interest in Fuzzy Logic Control and Machine Learning.
Lee Bissessar
Researcher
holds a BSc in ECNG from the University of the West Indies, Trinidad and Tobago. He led the creation of the Autonomous Systems MATLAB Simulation Package, enhancing system reliability using a model-driven approach. Lee is a primary developer at Intelligent Systems Labs, dedicated to dynamic simulation and innovative solutions in AI and robotics.
Derron Phagoo
Industry Partner
He is an engineer with a diverse background in software and hardware development across the Automotive, Commercial, and Oil & Gas industries, with expertise in compliance with Industry Standards. His interests include Artificial Intelligence, Automation, and Robotics.