Vision-driven autonomy for marine robotic systems
MARVIS Lab develops advanced control, computer vision, and AI methods for reliable underwater and maritime autonomy. We focus on visual servoing, model predictive control, and learning-enabled perception for AUVs, ROVs, and USVs operating in harsh, GNSS-denied marine environments.
Core Themes
- Marine vision & perception
- Vision-guided control & visual servoing
- Fault-tolerant & predictive control
- LLM-guided inspection & multimodal robotics
- Aquaculture & subsea infrastructure inspection
Research
MARVIS Lab sits at the intersection of marine robotics, computer vision, and intelligent control. We aim to bridge rigorous theory with field-deployable systems for sustainable ocean technologies.
Marine Vision & Perception
- Underwater image/video enhancement and understanding
- Semantic segmentation & detection for aquaculture and subsea assets
- Benchmarks for aquaculture net-pen inspection
Vision-Guided & Fault-Tolerant Control
- Visual servoing schemes for ROV/AUV inspection
- Model Predictive Control and set-theoretic methods
- Fault-tolerant thrust allocation and health-aware autonomy
Intelligent Autonomous Marine Systems
- Autonomous ROV, AUV & USV platforms
- UAV–USV cooperation for localization & guidance
- Mission planning for long-duration inspection
AI & LLM-guided Inspection
- AquaChat & AquaChat++ frameworks for ROV inspection
- Vision–Language–Action models for marine robotics
- Ontology-driven prompting and explainable decision making
People
Waseem Akram
Founder & Head of MARVIS Lab
Postdoctoral Fellow, Department of Mechanical Engineering, Khalifa University
Waseem Akram works on guidance and control of marine robotic systems, visual servoing, fault-tolerant control allocation, and learning-enabled perception for aquaculture and subsea inspection.
Students & Collaborators
- PhD, MSc, and undergraduate positions will be listed here.
- Current collaborations span Europe, the Middle East, and Asia.
Selected Publications
A short selection of recent work. For a complete list, please see Google Scholar or ResearchGate.
- Akram, W. et al., “A Visual Servoing Scheme for Autonomous Aquaculture Net Pens Inspection Using ROV,” Sensors, 2022.
- Tedesco, F., Akram, W. et al., “Predictive maintenance of actuators in linear systems: a receding horizon set-theoretic approach,” Int. J. Robust and Nonlinear Control, 2022.
- Akram, W. et al., “Aquaculture defects recognition via multi-scale semantic segmentation,” Expert Systems with Applications, 2023.
- Bakht, A. B. et al., Akram, W. co-author, “MuLA-GAN: Multi-level attention GAN for enhanced underwater visibility,” Ecological Informatics, 2024.
- Akram, W. et al., “Enhancing Aquaculture Net Pen Inspection: A Benchmark Study on Detection and Semantic Segmentation,” IEEE Access, 2025.
- Akram, W. et al., “AquaChat: An LLM-guided ROV framework for adaptive inspection of aquaculture net pens,” Aquacultural Engineering, 2025.
Contact
MARVIS Lab – Marine Vision & Intelligent Systems Laboratory
Department of Mechanical Engineering, Khalifa University
Abu Dhabi, United Arab Emirates
Email:
waseem.akram@dimes.unical.it
Personal page:
waseemlab.dotnest.com
Google Scholar: [link]
ResearchGate: [link]