A Vision for Inclusive, Interdisciplinary AI-based Collaborative Research
Title of the day:
As artificial intelligence revolutionizes the way we live, learn, and lead, human-driven cooperative innovation has never been so critical. In this context, LeadwitAI.net has been created to establish a platform that allows researchers, and innovators from all corners of the globe and across different disciplines to work together, exchange ideas, and co-create the future of smart systems. As an AI researcher and academic passionate about robotics, embedded systems, and neuroadaptive technologies, I have witnessed both the immense potential and the challenges of working in silos. Progress in AI and smart systems should not be confined to isolated labs or single-discipline approaches. That’s why this platform is dedicated to fostering collaborative research that is:
- Interdisciplinary — blending AI, neuroscience, embedded hardware, ethics, and more.
- International — connecting researchers and institutions across all regions.
- Inclusive — open to researchers, innovators, and underrepresented voices.
- Impact-Oriented — solutions that transform lives, not just publishing papers.
Through research briefs, calls to collaborate, opinion pieces, and blog entries, LeadwitAI.net will be an arena of knowledge sharing and voice for visionary, responsible AI leadership. If you’re a seasoned professor, an emerging scholar, or a policymaker with an interest in smart technologies, come walk this journey collaboratively, and boldly.
Research Directions for 2025: AI-Powered Innovation Across Domains:
Track 1: AI in Robotics & Cybersecurity
In this track, we explore the development of smart robotic systems for underwater and aerial applications, with a strong focus on security, robustness, and autonomous capabilities. Recent publications in this field:
- A New Design of a MEMS-Actuated Reconfigurable Soft Robotic Platform for Underwater Environmental and Security Applications, 2025
Under review; journal publication. - Hybrid Underwater Target Tracking with Particle Filters and Deep Learning, Focus on Design Considerations, 2024
- Design Considerations for Hybrid Underwater Target Tracking: Integrating Particle Filters and Deep Learning for Enhanced Decision, 2024
- Safeguarding a Flying Robot, Designed to Investigate the State of Smart Grid, against Cyberattacks, 2024
- A fuzzy logic-based stabilization system for a flying robot, with an embedded energy harvester and a visual decision-making system, 2023
- Blockchain-Based Heuristic Study to Secure UAVs From GPS Spoofing Signals and External Attacks, 2023
- A new design of a flying robot, with advanced computer vision techniques to perform self-maintenance of smart grids, 2022
Track 2: AI in Medical and Neuroscience Applications
This research track aims to bridge artificial intelligence and neuroscience for improved diagnostics, therapy, and disease management. Collaboration with neuroscience experts is actively sought. Recent publications in this field:
- Enhancing Neural Rehabilitation Insights: On the Path of Bridging Artificial and Biological Neural Networks, 2024
- A New Technique Integrating Adam Optimizer and Q-Learning Algorithm for Adaptive Neurofeedback Therapy, 2024
- Neural Networks from Biological to Artificial and Vice Versa, 2024
Current Collaborative Projects:
A collaborative project between the Gulf University for Science and Technology, Kuwait College of Science and Technology, and the University of Sialkot, titled “Web-based, AI-driven predictive platform for disease management” has been established. This project is supported by the Seed Grant from Gulf University for Science and Technology – Graduate Studies and Research, under the umbrella of the Center for Applied Mathematics & Bioinformatics (CAMB), 2024.
A New Proposal:
Neuroscientists with expertise in Dopaminergic systems, Clinical neurorehabilitation, and Neuroimaging are invited to collaborate in the following research project that merges mechanistic neuroscience with cutting-edge AI, offering a novel pathway to personalized neurorehabilitation. We need to gain a computational tool to test hypotheses about dopamine’s role in learning, and generate a ground algorithm in biological reality, enhancing clinical applicability.
Project title
Bridging Neurochemistry and Machine Learning: A Dopamine-Inspired Adaptive Optimizer for Personalized Neurorehabilitation
Abstract:
Recent improvements in artificial neural networks (ANNs) and spiking neural networks (SNNs) have shown great promise in modeling neuropsychological disorders and optimizing neurofeedback treatments. Still, none of the existing optimization algorithms, for example, Adam or Q-learning, contain the biological fidelity required for valid modeling of neuromodulator dopaminergic reinforcement learning in the human brain, being a case in point.
This project is seeking to:
· Construct an optimizer inspired by dopamine that dynamically modifies learning rates and synaptic plasticity by considering reward signals like biological reinforcement mechanisms.
· Validate the optimizer in neurorehabilitation tasks, for example, stroke recovery and Parkinson’s disease, with EEG/fMRI-guided neurofeedback.
· Design a closed-loop and adaptive neurorehabilitation system that provides real-time personalized therapy.
Neuroscience Potential Contribution Vs AI Contribution
· Provide data on dopamine’s role in reward-based plasticity
Design an optimizer where learning rates adapt via simulated dopamine signals
· Define clinically relevant EEG/fMRI biomarkers for reinforcement
Integrate the optimizer with deep reinforcement learning for dynamic parameter tuning.
· Co-design experiments with patient data
Test the system in silico before pilot trials.
Expected Outcomes
· A biologically plausible optimizer that would perform better in neurorehabilitation tasks than standard Adam/Q-learning algorithms.
· Proof-of-concept neurofeedback enabling dynamic personalization.
Potential grants:
NIH BRAIN Initiative, Wellcome Trust, or ERC Horizon Europe.
A Selection of Recent Publications:
- A new design of a flying robot, with advanced computer vision techniques to perform self-maintenance of smart grids
- Advanced AI-based techniques to predict daily energy consumption: A case study
- Electricity-consuming forecasting by using a self-tuned ANN-based adaptable predictor
- FPGA-based parallel implementation to classify Hyperspectral images by using a Convolutional Neural Network
- Neural networks from biological to artificial and vice versa
