About

Giovanni De Magistris is a leading researcher at the intersection of robotics, deep reinforcement learning (DRL), and industrial automation. His most impactful work demonstrates how robots can overcome physical precision limits using intelligent algorithms. His seminal paper, "Deep Reinforcement Learning for High Precision Assembly Tasks" (300 citations), pioneered the use of DRL to enable robots to perform tight-clearance peg-in-hole tasks without tedious manual parameter tuning—a breakthrough for manufacturing. To make DRL safe for real-world deployment, he introduced "OptLayer" (118 citations), a constrained optimization framework that prevents unsafe trial-and-error interactions. De Magistris has also advanced sim-to-real transfer learning, using variational autoencoders to bridge the gap between synthetic training data and real-world vision (38 citations), and developed unsupervised anomaly detection systems for industrial robots (45 citations). His work extends beyond factories to underwater surveillance, where he has contributed to cooperative autonomy for networked AUVs. With over 600 total citations, De Magistris’s research is essential reading for anyone working on practical, safe, and precise robotic manipulation.

Research Focus

Key Achievements

11
H-Index
20
Papers
685
Total Citations
34
Avg Citations/Paper
🏆 Most Cited Paper
Deep reinforcement learning for high precision assembly tasks
300 citations · 2017
📈 Most Prolific Year: 2017 (8 Papers)
🤝 Key Collaborators: 45
🏛 Institutions: IBM Research - Tokyo, École Normale Supérieure - PSL, North Atlantic Treaty Organization, Centre National de la Recherche Scientifique, National Institute of Advanced Industrial Science and Technology

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago