Research

Task and Motion Planning

We develop algorithms that enable robots to autonomously plan and execute complex tasks by coordinating high-level task reasoning with low-level motion control. These methods allow robots to operate robustly in dynamic and unstructured environments.

Learning-Based Control

Our learning-based control approaches leverage machine learning to improve robot performance through experience. By adapting to variability and uncertainty, these systems enable more flexible and resilient robotic behavior.

Adaptive Autonomy

We study adaptive autonomy to dynamically balance human involvement and robot independence. This allows robotic systems to adjust their level of autonomy based on task complexity, user intent, and environmental conditions.

Enhanced Surgical Robotics Training

We introduced Haptic Assistance algorithms into Virtual Reality environments used in the context of training robotics surgeons.

With an assisted training protocol, surgeons will undergo an impreved learning pathway that improves surgical performance and favors skill retention

Human-robot skills transfer

To release the complexity of surgical operations and reduce the workloads of surgeons, human-robot skill transferring is proposed to improve the practicability of surgical robots.

Learning the surgical skills from surgeon experts and reproduce these surgical operations is an effective way to enhance the autonomy of the surgical robot in RAMIS. Three stages are included in this process, namely, demonstration, learning, and reproduction.

Human-Robot shared control for Ultrasound Scanning

Ultrasound scanning benefits from a robotized approach in terms of reaching optimal resolution, 3D reconstruction and optimized force control

We embed an ultrasound probe on a robotic arm to execute a scanning procedure more effectively and with better results.

AR Surgical Navigation

We develop augmented reality systems for surgical navigation that overlay critical information directly onto the clinician’s view. This enhances spatial awareness, precision, and decision-making during complex procedures.

Intraoperative Guidance

Our AR-based intraoperative guidance solutions provide real-time visual cues and feedback during interventions. By integrating imaging, tracking, and patient-specific data, these systems support safer and more effective surgical workflows.

AR User Interface Design

We design AR user interfaces that are intuitive, ergonomic, and tailored to the needs of clinicians. Our focus is on minimizing cognitive load while maximizing clarity and interaction efficiency in high-stakes environments.

Federated Learning

We apply federated learning techniques to enable collaborative model training across distributed clinical data sources while preserving data privacy. This approach allows robust and scalable learning without sharing sensitive patient data.

Explainable AI for Diagnosis

We develop explainable AI methods that provide transparent and interpretable insights into diagnostic decisions. These techniques support clinical trust, regulatory compliance, and informed decision-making by healthcare professionals.

Data-Driven Prediction of Clinical Events

Our data-driven models analyze multimodal clinical data to predict adverse events and disease progression. By enabling early detection and personalized risk assessment, these approaches support proactive and preventive healthcare