Why It Matters
Understanding how robots can sense and adapt in real time is crucial for scaling robotics into healthcare and sustainable manufacturing, where precision and safety are paramount. As the push for automation and recycling intensifies, these advances promise to reduce human fatigue, improve surgical outcomes, and make electronic waste processing more efficient, making the episode highly relevant to both industry professionals and the broader public.
Key Takeaways
- •Robots combine vision, force sensing for real-time manipulation.
- •Dexterous grasping enables electronic disassembly and recycling.
- •Medical robots use ultrasound and haptics for needle procedures.
- •Autonomy in surgery remains low; focus on collaborative assistance.
- •Learning from demonstration mirrors human infant skill acquisition.
Pulse Analysis
The Autonomous Robotic Manipulation Lab at Heriot‑Watt treats sensing as a continuous feedback loop, not a pre‑step. By fusing computer vision, force sensing and machine‑learning, robots grasp delicate, irregular objects with human‑like dexterity. This interdisciplinary mix of hardware design, AI and human‑robot interaction yields prototypes that assemble and disassemble complex electronics—tasks traditionally done by skilled technicians. Real‑time perception lets robots adapt on the fly, turning unstructured environments into manageable workspaces. The team also explores bio‑inspired learning, allowing robots to refine grasps through trial and error.
In medical robotics the same perception‑action loop powers autonomous needle procedures such as IV access and thoracentesis. Ultrasound imaging combined with haptic feedback tracks organ motion and adjusts needle trajectory for sub‑millimetre accuracy. Full autonomy remains limited; systems operate under high human supervision, augmenting surgeons with steadier hands and less fatigue. Researchers explore autonomy levels from tele‑operated assistance to semi‑autonomous subtasks like vein detection and autonomous insertion, showing how collaborative robots can improve safety and broaden access to advanced care. Such feedback loops also enable rapid adaptation when patients breathe or shift, maintaining target alignment.
The lab also applies autonomous manipulation to waste recycling. Robotic arms equipped with suction grippers and multispectral vision classify items on fast conveyor belts, handling occlusions and varied materials. AI models trained via learning‑from‑demonstration enable robots to separate tiny screws or fragile components, reducing human exposure to hazardous conditions. These systems increase sorting efficiency and lower labor costs, supporting circular‑economy goals. Future deployments aim to integrate real‑time material analytics, further improving sorting precision and resource recovery. As companies scale such technology, the blend of perception, dexterity and machine learning will become essential for greener manufacturing and higher recycling rates.
Episode Description
Maria Koskinopoulou is an Assistant Professor in Robotics and Computer Vision at Heriot-Watt University. She co-leads the ARM²Lab – Autonomous Robotic Manipulation & Multi-Agent Systems Lab at Heriot-Watt and the National Robotarium, alongside Ignacio Carlucho. Her research interests include robotic manipulation, perception, robot vision, medical robotics, human-robot interaction, and machine learning. She is involved in major UKRI and EU-funded research projects advancing robotic manipulation, surgical and underwater robotics, autonomous assembly, and waste sorting.
Check out the bonus episodes from the European Robotics Forum: https://www.patreon.com/posts/robot-talk-at-157276631
Join the Robot Talk community on Patreon: https://www.patreon.com/ClaireAsher
Comments
Want to join the conversation?
Loading comments...