Robotics, Embodied AI, Navigation in vivo

Thrilled to share our latest work, 𝐆𝐞𝐨𝐋𝐚𝐧𝐆, a unified geometry-aware framework for language-guided robotic grasping.

Language-guided grasping is a key capability for intuitive human–robot interaction. A robot should not only detect objects but also understand natural instructions such as “pick up the blue cup behind the bowl.” While recent multimodal models have shown promising results, most existing approaches rely on multi-stage pipelines that loosely couple perception and grasp prediction. These methods often overlook the tight integration of geometry, language, and visual reasoning, making them fragile in cluttered, occluded, or low-texture environments. This motivated us to bridge the gap between semantic language understanding and precise geometric grasp execution.

🧠✨ 𝐖𝐡𝐚𝐭 𝐰𝐞 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐝:

A novel unified framework for geometry-aware language-guided grasping that includes:

🔹 Unified RGB-D Multimodal Representation:

 We embed RGB, depth, and language features into a shared representation space, enabling consistent cross-modal semantic alignment for accurate target reasoning.

🔹 Depth-Guided Geometric Module (DGGM):

 Instead of treating depth as auxiliary input, we explicitly inject geometric priors derived from depth into the attention mechanism, strengthening object discrimination under occlusion and ambiguous visual conditions.

🔹 Adaptive Dense Channel Integration (ADCI):

 A dynamic multi-layer fusion strategy that balances global semantic cues and fine-grained geometric details for robust grasp prediction.

🎯  𝐊𝐞𝐲 𝐑𝐞𝐬𝐮𝐥𝐭𝐬:

✅ GeoLanG significantly outperforms prior multi-stage baselines on OCID-VLG for language-guided grasping.

✅ Demonstrates strong robustness in cluttered and heavily occluded scenes.

✅ Successfully validated on real robotic hardware, showing reliable sim-to-real transfer.

💡 𝐖𝐡𝐲 𝐢𝐭 𝐦𝐚𝐭𝐭𝐞𝐫𝐬:

This work shows that tightly coupling geometric reasoning with multimodal language understanding can significantly enhance robotic grasp reliability. By embedding depth-aware geometric priors directly into attention mechanisms, we reduce ambiguity and improve consistency in grasp decision-making.

GeoLanG provides a pathway toward more intelligent robotic systems that understand not just what object to grasp, but also how to grasp it robustly in complex real-world environments.

🌱 𝐖𝐡𝐚𝐭’𝐬 𝐧𝐞𝐱𝐭?

We are exploring extending this geometry-aware multimodal reasoning toward:

 🔹 Real-time interactive grasping

 🔹 Multi-step manipulation tasks

 🔹 Integration with motion planning and autonomous robotic control

#ICRA2026 #CUHK

Thrilled to share our latest work on enabling robust sparse-to-dense reconstruction for endoscopic surgical robots — bridging the gap between 𝐬𝐩𝐚𝐫𝐬𝐞 𝐬𝐞𝐧𝐬𝐨𝐫 𝐝𝐚𝐭𝐚 𝐚𝐧𝐝 𝐡𝐢𝐠𝐡-𝐪𝐮𝐚𝐥𝐢𝐭𝐲 𝟑𝐃 𝐦𝐚𝐩𝐩𝐢𝐧𝐠 using a novel 𝐝𝐢𝐟𝐟𝐮𝐬𝐢𝐨𝐧-𝐛𝐚𝐬𝐞𝐝 framework.

Fine-tuning foundational models often fails due to a lack of dense ground truth, and self-supervised methods struggle with scale ambiguity, sparse depth sensors offer a reliable geometric prior.

This motivated us to develop EndoDDC, a method that robustly generates dense depth maps by fusing RGB images with sparse depth inputs.

🧠✨ 𝐖𝐡𝐚𝐭 𝐰𝐞 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐝:

A diffusion-driven depth completion architecture that:

🔹 Integrates sparse depth and RGB inputs to overcome the limitations of pure visual estimation.

🔹 Utilizes a Multi-scale Feature Extraction and Depth Gradient Fusion module to capture fine-grained surface orientation and local structure.

🔹 Optimizes depth maps iteratively using a conditional diffusion model, refining geometry even in regions with weak textures or reflections.

🎯 𝐊𝐞𝐲 𝐑𝐞𝐬𝐮𝐥𝐭𝐬:

✅ 25.55% and 9.03% improvement in accuracy on the StereoMIS and C3VD dataset compared to SOTA surgical estimators like EndoDAC.

✅ 7.35% and 5.28% reduction in RMSE on StereoMIS and C3VD compared to the best depth completion baseline (OGNI-DC).

✅ Outperformed foundational models (DepthAnything-v2) and standard depth completion (Marigold-DC) methods in both accuracy and robustness.

💡 𝐖𝐡𝐲 𝐢𝐭 𝐦𝐚𝐭𝐭𝐞𝐫𝐬:

This work demonstrates that diffusion models can effectively solve the “sparse-to-dense” challenge in medical imaging. By providing accurate depth completion despite complex lighting and texture conditions, EndoDDC has the potential to significantly enhance autonomous navigation, procedural safety, and spatial awareness in minimally invasive surgery.

🔖 #DepthCompletion #DiffusionModel #EndoscopicSurgery #SurgicalNavigation #ICRA #CUHKEngineering #CUHK

We present 𝐍𝐞𝐮𝐫𝐨-𝐕𝐋𝐀, an scenario-aware model designed for the motion control of a parallel continuum neurosurgical robot.

Robotic surgery systems have garnered significant attention for their precision and efficiency, yet achieving autonomous tasks in complex neurosurgical environments remains challenging. Although Vision-Language-Action (VLA) models hold great potential, their development is constrained by the scarcity of data from surgical environments and robotic kinematics. To address this issue, this paper proposes NeuroVLA: a VLA model specifically designed for neurosurgical robotic tumor debulking tasks. Through phantom experiments conducted on a flexible parallel continuum robot, we constructed a dataset and decomposed the debulking task into four skill-based instructions. NeuroVLA utilizes a Vision-Language Model (VLM) as its backbone for scene reasoning, enabling the robot to comprehend the surgical scene and its own state. Experimental results demonstrate that after training on 90 debulking segments, NeuroVLA can infer actions based on images, language instructions, and the robot’s state. It achieved average pixel distance errors of 29.10 pixels and 21.55 pixels for the “alignment” and “transfer” skills, respectively, and success rates of 88.89% and 100% for the “grasping” and “release” skills.

🧠 Technical Framework:

●     End-to-End scenario-aware VLA model

●     Skill-based scenario infer mechanism

●     Debulking task dataset in neurosurgery

🎯 Experimental Results:

●     NeuroVLA demonstrates significantly lower pixel distance (PD) errors in the “alignment” and “transfer” skills (29.10 px / 21.55 px), far surpassing the performance of baseline models (such as Octo’s 79.72 px / 65.46 px).

In the “grasping” and “release” skills, NeuroVLA exhibits greater robustness, achieving a grasping success rate of 88.89% and a release success rate of 100%. In contrast, baseline models often misinterpret incomplete forceps closure as task completion, leading to grasping failures.

#ICRA2026

We present 𝐊𝐢𝐫𝐢-𝐂𝐚𝐩𝐬𝐮𝐥𝐞, a swallowable kirigami-inspired capsule robot that enables minimally invasive GI biopsy—pushing capsule endoscopy from imaging to tissue sampling.

Wireless capsule endoscopy is comfortable and accessible, but cannot collect biopsy tissue, while histology is still the gold standard. Our work targets safe, depth-controlled, retrievable sampling in a capsule form factor.

🧠 Technical Framework:

●     Kirigami PI skin: flat during locomotion, deploys sharp protrusions when stretched

●     Dual-cam actuation: compact, repeatable deployment + recovery

●     Rotary scraping + internal storage: detach tissue and store it in internal cavities for retrieval

🎯 Experimental Results:

●     Penetration depth: median ~0.61 mm (0.46–0.66 mm) on ex vivo porcine tissue

●     Biopsy yield: ~10.9 mg (stomach) / 18.9 mg (small intestine), histology-ready (mucosa + submucosa)

●     Forces within reported GI biopsy safety ranges

💡 Significance:

A practical route toward capsule-based biopsy with controlled shallow interaction and retrievable specimens—without bulky endoscopic tools.

🌱 Next:

Moving toward untethered actuation (e.g., magnetic drive) and multi-site sampling to reduce mixing risk.

🔖 #ICRA2026 #SoftRobotics #MedicalRobotics #CapsuleEndoscopy #Kirigami #BioInspiredRobotics #MinimallyInvasive #Biopsy #GI #RoboticsResearch

𝐔𝐧𝐝𝐞𝐫𝐬𝐭𝐚𝐧𝐝𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝐋𝐚𝐫𝐠𝐞 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐌𝐨𝐝𝐞𝐥 𝐢𝐧 𝐑𝐨𝐛𝐨𝐭-𝐚𝐬𝐬𝐢𝐬𝐭𝐞𝐝 𝐒𝐮𝐫𝐠𝐞𝐫𝐲!

Thrilled to share our latest work, 𝐒𝐮𝐫𝐠𝐕𝐢𝐝𝐋𝐌, the first video-language model specifically designed to address both full and fine-grained surgical video comprehension.

Surgical scene understanding is critical for training and robotic decision-making. While current Multimodal Large Language Models (MLLMs) excel at image analysis, they often overlook the fine-grained temporal reasoning required to capture detailed task execution and specific procedural processes within a surgery. This motivated us to bridge the gap between global video understanding and micro-action analysis.

🧠✨ What we developed:

A novel framework and resource for surgical video reasoning that includes:

🔹 𝐓𝐰𝐨-𝐬𝐭𝐚𝐠𝐞 𝐒𝐭𝐚𝐠𝐞𝐅𝐨𝐜𝐮𝐬 𝐦𝐞𝐜𝐡𝐚𝐧𝐢𝐬𝐦: The first stage extracts global procedural context, while the second stage performs high-frequency local analysis for fine-grained task execution.

🔹 𝐌𝐮𝐥𝐭𝐢-𝐟𝐫𝐞𝐪𝐮𝐞𝐧𝐜𝐲 𝐅𝐮𝐬𝐢𝐨𝐧 𝐀𝐭𝐭𝐞𝐧𝐭𝐢𝐨𝐧 (𝐌𝐅𝐀): Effectively integrates low-frequency global features with high-frequency local details to ensure comprehensive scene perception.

🔹 𝐒𝐕𝐔-𝟑𝟏𝐊 𝐃𝐚𝐭𝐚𝐬𝐞𝐭: We constructed a large-scale dataset with over 31,000 video-instruction pairs, featuring hierarchical knowledge representation for enhanced visual reasoning.

🎯 Key Results:

✅ SurgVidLM significantly outperforms existing models (like Qwen2-VL) in multi-grained surgical video understanding tasks.

✅ Capable of inferring anatomical landmarks (e.g., Denonvilliers’ fascia) and providing clinical motivation, moving beyond simple visual description.

✅ Demonstrated strong performance on unseen surgical tasks, proving the robustness of our hierarchical training approach.

💡 Why it matters:

This work shows that by combining global context with localized high-frequency focus, we can significantly reduce “hallucinations” in surgical AI. It provides a pathway toward more intelligent, context-aware surgical assistants that can understand not just what is happening, but how and why specific steps are performed.

🌱 What’s next?

We are exploring how to extend this multi-grained understanding to real-time intraoperative guidance and integrating it with physical robotic control for autonomous sub-tasks.

We present TMR-VLA, an end-to-end framework designed for the motion control of tri-leg silicone-based soft robots.

Miniature magnetic robots face a hardware bottleneck where the robot body is too small to integrate onboard sensors or power. This creates a gap between actuation and perception, often requiring human experts to manually adjust magnetic fields based on visual feedback. Our work aims to bridge this gap by enabling autonomous control through a multi-modal system.

🧠 Technical Framework:

●     End-to-End Mapping: The policy translates sequential endoscope images and natural language instructions directly into low-level coil voltage commands.

●     Action Adaptor: We utilized an EndoVLA-initialized backbone with an Action LoRA Adaptor that allows the model to autoregressively emit voltage increments.

●     TrilegMR-Motion Dataset: The model was trained on a new dataset containing 15,793 image-action pairs across 60 episodes.

●     Diverse Locomotion: The system controls five motion primitives: squatting, leg-lifting, rotation, forward movement, and recovery.

🎯 Experimental Results:

●     Success Rate: TMR-VLA achieved an average success rate of 74% across tested motion types.

●     Performance: The model outperformed general-purpose multimodal models (such as Qwen2.5-VL and LLaVA-1.6) in both instruction interpretation and action execution.

●     Inference Speed: Real-time control was demonstrated at approximately 2 Hz using an NVIDIA RTX 5090 GPU.

💡 Significance: This study addresses the challenge of autonomous control in untethered soft robots without increasing their structural complexity. It provides a foundational baseline for intelligent navigation in complex in-vivo environments.

Big news! 🚀 Our lab is proud to announce that 6 of our latest papers have been accepted! 🎊

We are incredibly proud of the team’s hard work and innovation. To give each project the spotlight it deserves, we will be sharing details about each breakthrough one by one over the coming days.

Stay tuned for the updates! 🤖✨

We are thrilled to share our latest Comment published in #NatureReviewsBioengineering: “𝗔𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗸𝗶𝗻𝗮𝗲𝘀𝘁𝗵𝗲𝘀𝗶𝗮 𝗶𝗻 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗿𝗼𝗯𝗼𝘁𝗶𝗰 𝘀𝘂𝗿𝗴𝗲𝗿𝘆”.

Current autonomous surgical robots are 𝗵𝗲𝗮𝘃𝗶𝗹𝘆 𝘃𝗶𝘀𝗶𝗼𝗻-𝗰𝗲𝗻𝘁𝗿𝗶𝗰. While they 𝗰𝗮𝗻 “𝘀𝗲𝗲” anatomy, they 𝗹𝗮𝗰𝗸 𝘁𝗵𝗲 𝗶𝗻𝘁𝗿𝗶𝗻𝘀𝗶𝗰 𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝘁𝗼 “𝗳𝗲𝗲𝗹” tissue interactions—𝗮 𝗰𝗿𝘂𝗰𝗶𝗮𝗹 𝘀𝗸𝗶𝗹𝗹 𝘁𝗵𝗮𝘁 𝗵𝘂𝗺𝗮𝗻 𝘀𝘂𝗿𝗴𝗲𝗼𝗻𝘀 𝗿𝗲𝗹𝘆 𝗼𝗻 𝗳𝗼𝗿 𝘀𝗮𝗳𝗲𝘁𝘆 𝗮𝗻𝗱 𝗱𝗲𝘅𝘁𝗲𝗿𝗶𝘁𝘆.

In this article, we propose a hierarchical framework for 𝗔𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗞𝗶𝗻𝗮𝗲𝘀𝘁𝗵𝗲𝘀𝗶𝗮 to bridge this gap:

1. 📈𝗧𝗵𝗲 𝗣𝗵𝘆𝘀𝗶𝗰𝗮𝗹 𝗟𝗲𝘃𝗲𝗹: Integrating proprioception and exteroception for high-res physical sensing.

2. 💬𝗧𝗵𝗲 𝗔𝗹𝗴𝗼𝗿𝗶𝘁𝗵𝗺𝗶𝗰 𝗟𝗲𝘃𝗲𝗹: Moving from raw signal processing to semantic understanding of contact.

3. 🧠𝗧𝗵𝗲 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗮𝗹 𝗟𝗲𝘃𝗲𝗹: Implementing Vision-Kinaesthesia-Language-Action models to achieve true sensorimotor synergy.

We believe the future of autonomous surgery lies in systems that can synergistically fuse vision and kinaesthesia to not just see, but truly feel, think, and act.

📃 Read the 𝗳𝘂𝗹𝗹 𝗽𝗮𝗽𝗲𝗿 here: [https://lnkd.in/gqTEpYjs]

👏 Kudos to our amazing team, Dr. Tangyou Liu, Dr. Sishen YUAN, and Prof. Hongliang Ren.

Thrilled to share our latest The International Journal of Robotics Research (IJRR) work on enabling 𝗴𝗿𝗮𝘃𝗶𝘁𝘆‑𝗮𝘄𝗮𝗿𝗲 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 for 𝗽𝗼𝗿𝘁𝗮𝗯𝗹𝗲 𝗰𝗮𝗯𝗹𝗲‑𝗱𝗿𝗶𝘃𝗲𝗻 𝘀𝗼𝗳𝘁 𝘀𝗹𝗲𝗻𝗱𝗲𝗿 𝗿𝗼𝗯𝗼𝘁𝘀 — using 𝗼𝗻𝗹𝘆 𝗮 𝘀𝗶𝗻𝗴𝗹𝗲 𝗜𝗠𝗨 and a powerful 𝗿𝗼𝗯𝗼𝗽𝗵𝘆𝘀𝗶𝗰𝗮𝗹 simulation‑driven framework.

Soft robots are lightweight and flexible, but their high aspect ratios make them 𝘦𝘹𝘵𝘳𝘦𝘮𝘦𝘭𝘺 sensitive to gravity, causing passive deformation that traditional kinematics just can’t handle. This motivated us to rethink how soft robots can 𝘴𝘦𝘯𝘴𝘦 and 𝘤𝘰𝘮𝘱𝘦𝘯𝘴𝘢𝘵𝘦 for gravity — without bulky sensors or complex hardware.

🧠✨ 𝗪𝗵𝗮𝘁 𝘄𝗲 𝗱𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗱:

A 𝗯𝗶𝗼‑𝗶𝗻𝘀𝗽𝗶𝗿𝗲𝗱 𝗿𝗲𝗮𝗹𝟮𝘀𝗶𝗺𝟮𝗿𝗲𝗮𝗹 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 that:

🔹 Streams real‑world IMU orientation into a real‑time and robust SOFA simulation

🔹 Dynamically reorients virtual gravity to mirror reality

🔹 Uses QP optimization to compute joint‑level compensation

🔹 Executes compensation in both simulation and the physical robot

All of this — using just 𝗼𝗻𝗲 𝗜𝗠𝗨. No strain sensors. No cameras. No expensive reconstruction systems.

🎯 𝗞𝗲𝘆 𝗥𝗲𝘀𝘂𝗹𝘁𝘀:

✅ >99% compensation recovery in static tests

✅ ~94% recovery in low‑motion dynamic tests

✅ Demonstrated on different two-segment cable‑driven soft robots

💡 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀:

This work shows that soft robots can maintain 𝘀𝘁𝗮𝗯𝗹𝗲, 𝗰𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝘁 𝗰𝗼𝗻𝗳𝗶𝗴𝘂𝗿𝗮𝘁𝗶𝗼𝗻𝘀 𝘂𝗻𝗱𝗲𝗿 𝗰𝗵𝗮𝗻𝗴𝗶𝗻𝗴 𝗴𝗿𝗮𝘃𝗶𝘁𝘆 by integrating 𝘃𝗶𝗿𝘁𝘂𝗮𝗹 𝘀𝗲𝗻𝘀𝗶𝗻𝗴 + 𝘀𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻‑𝗱𝗿𝗶𝘃𝗲𝗻 𝗶𝗻𝘃𝗲𝗿𝘀𝗲 𝗰𝗼𝗺𝗽𝘂𝘁𝗮𝘁𝗶𝗼𝗻. It reduces reliance on physical sensors and opens a pathway toward 𝘀𝗰𝗮𝗹𝗮𝗯𝗹𝗲, 𝗴𝗲𝗻𝗲𝗿𝗮𝗹𝗶𝘇𝗮𝗯𝗹𝗲 𝗴𝗿𝗮𝘃𝗶𝘁𝘆‑𝗮𝘄𝗮𝗿𝗲 𝘀𝗼𝗳𝘁 𝗿𝗼𝗯𝗼𝘁𝘀.

🌱 𝗪𝗵𝗮𝘁’𝘀 𝗻𝗲𝘅𝘁?

We’re exploring how to extend this architecture to virtualizable external force sensing and richer environmental interactions.

Special shoutout to the team —

Jiewen Lai, Tian-Ao Ren (co‑first authors),

Pengfei YE, Yanjun Liu, Jingyao Sun, Hongliang Ren —

for making this project possible.

🔗 Paper link: https://lnkd.in/gMgwCfvf

𝗧𝗶𝘁𝗹𝗲: 𝗔 𝗧𝗿𝗮𝗻𝘀𝗲𝗻𝗱𝗼𝘀𝗰𝗼𝗽𝗶𝗰 𝗧𝗲𝗹𝗲𝗿𝗼𝗯𝗼𝘁𝗶𝗰 𝗦𝘆𝘀𝘁𝗲𝗺 𝗨𝘀𝗶𝗻𝗴 𝗛𝗲𝘁𝗲𝗿𝗼𝗴𝗲𝗻𝗲𝗼𝘂𝘀 𝗙𝗹𝗲𝘅𝗶𝗯𝗹𝗲 𝗠𝗮𝗻𝗶𝗽𝘂𝗹𝗮𝘁𝗼𝗿𝘀 𝗳𝗼𝗿 𝗕𝗶𝗺𝗮𝗻𝘂𝗮𝗹 𝗘𝗻𝗱𝗼𝘀𝗰𝗼𝗽𝗶𝗰 𝗦𝘂𝗯𝗺𝘂𝗰𝗼𝘀𝗮𝗹 𝗗𝗶𝘀𝘀𝗲𝗰𝘁𝗶𝗼𝗻

🔍 𝗕𝗮𝗰𝗸𝗴𝗿𝗼𝘂𝗻𝗱:

Endoscopic submucosal dissection (ESD) is a key technique for early GI cancer treatment, requiring high dexterity and precision.

🛠 𝗪𝗵𝗮𝘁 𝘄𝗲 𝗱𝗶𝗱:

We developed the first 𝗵𝗲𝘁𝗲𝗿𝗼𝗴𝗲𝗻𝗲𝗼𝘂𝘀 𝗳𝗹𝗲𝘅𝗶𝗯𝗹𝗲 𝗺𝗮𝗻𝗶𝗽𝘂𝗹𝗮𝘁𝗼𝗿𝘀 (𝗛𝗙𝗠𝘀) for bimanual ESD, integrating:

🤖 𝗦𝗲𝗿𝗶𝗮𝗹 𝗔𝗿𝘁𝗶𝗰𝘂𝗹𝗮𝘁𝗲𝗱 𝗠𝗮𝗻𝗶𝗽𝘂𝗹𝗮𝘁𝗼𝗿 (𝗦𝗔𝗠) – for stable, multidirectional tissue traction

🔬 𝗣𝗮𝗿𝗮𝗹𝗹𝗲𝗹 𝗖𝗼𝗻𝘁𝗶𝗻𝘂𝘂𝗺 𝗪𝗿𝗶𝘀𝘁 (𝗣𝗖𝗪) – for accurate tissue dissection

📐 𝗞𝗲𝘆 𝗰𝗼𝗻𝘁𝗿𝗶𝗯𝘂𝘁𝗶𝗼𝗻𝘀:

✔ Kinematic modeling using Denavit–Hartenberg & Cosserat rod methods

✔ Workspace & dexterity analysis via simulation

✔ Validation through 16 ex vivo ESD tests

💡 This work demonstrates a novel strategy for surgical robotics—leveraging heterogeneous structures to enhance flexibility, stiffness, and accuracy in minimally invasive procedures.

👏 Kudos to our amazing team and collaborators from CUHK (Prof. Huxin Gao, Tao Zhang, Prof. Hongliang Ren), Qilu Hospital (Xiaoxiao Yang, Prof. 左秀丽, Prof. Yanqing Li), Southern University of Science and Technology (Xiao Xiao, Prof. Qinghu Meng), and Beijing Institute of Technology (Prof. Changsheng Li)!

📖 Stay tuned for the full article in IEEE RAM!

🏆 Paper: Contact-Aided Navigation of Flexible Robotic Endoscope Using Deep Reinforcement Learning in Dynamic Stomach

👩‍🔬 Authors: Chi Kit Ng, Huxin Gao, Tianao Ren, Prof. Jiewen Lai, and Prof. Hongliang Ren

🔍 𝗪𝗵𝘆 𝗶𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀:

Navigating flexible robotic endoscopes in the dynamic, deformable stomach environment is a grand challenge. Our proposed Contact-Aided Navigation (CAN) strategy, powered by deep reinforcement learning and force-feedback, achieved:

• 100% success rate in both static and dynamic simulated stomach environments

• Average navigation error of just 𝟭.𝟲 𝗺𝗺

• Robust generalization even under strong external disturbances

This work highlights how 𝗲𝗺𝗯𝗼𝗱𝗶𝗲𝗱 𝗔𝗜 𝗮𝗻𝗱 𝗯𝗶𝗼𝗺𝗲𝗰𝗵𝗮𝗻𝗶𝗰𝘀-𝗶𝗻𝘀𝗽𝗶𝗿𝗲𝗱 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 can transform surgical robotics, enabling safer and more precise navigation in complex clinical environments.

Check the paper at https://lnkd.in/g6KgZTdD

🙏 Huge thanks to the team, collaborators, and the broader robotics community for the support and inspiration.