Papers

8

Total Citations

161

H-Index

5

About

Marios Kiatos is a leading researcher in robotic manipulation, specializing in autonomous grasping and object singulation in highly cluttered environments. His work addresses one of robotics’ most persistent challenges: reliably extracting a target object from a pile of others without collisions. Kiatos’s key contributions include developing a geometric approach for grasping unknown objects with multifingered hands, which leverages point cloud data to plan stable grasps without prior object models. His seminal paper, “Robust Object Grasping in Clutter via Singulation” (53 citations), introduced a strategy that separates target objects from surrounding clutter before grasping, dramatically improving success rates. Kiatos further advanced the field with modular reinforcement learning for total singulation and a push-grasping policy that uses non-prehensile actions to create space for multifingered hands. His work on split deep Q-learning and pre-grasp manipulation policies has pushed the boundaries of what robots can achieve in dense clutter. With over 160 total citations, Kiatos’s research is foundational for applications in warehouse automation, domestic robotics, and industrial sorting, where robots must operate reliably in unstructured, crowded spaces.

Research Focus

Key Achievements

5
H-Index
8
Papers
161
Total Citations
20
Avg Citations/Paper
🏆 Most Cited Paper
Robust object grasping in clutter via singulation
53 citations · 2019
📈 Most Prolific Year: 2019 (2 Papers)
🤝 Key Collaborators: 8
🏛 Institutions: Aristotle University of Thessaloniki, Centre for Research and Technology Hellas

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago