Home
Portrait of Javier Felip León

Dr. Javier Felip León

Staff AI / Robotics Scientist (Ph.D.)
Contact-rich, data-driven robotic manipulation | Reactive primitives | Sim-to-Real | Human intent prediction | High-DOF platforms | 50+ patents | 20+ publications
LinkedIn
CV (PDF)
PhD Thesis
Scholar
Patents
Email
Phone
Summary Experience Education Projects Skills Awards Publications Patents

Summary

Staff AI/Robotics Scientist (Ph.D.) with +18 years experience in robotics across academia and industry. Specializing in contact-rich, data-driven manipulation: built reactive primitive frameworks, sim-to-real and synthetic data pipelines, intuitive physics prediction, and human intent inference on high-DOF (dual-arm PA-10, UR, Fanuc, Baxter) platforms; 50+ granted patents and 20+ publications underpin robust grasp planning, perception, and autonomy stack contributions. Proficient in Python/C++, Linux, ROS2, PyTorch, MoveIt!, PyBullet/Drake/Gazebo and Blender.

Work Experience

Staff AI/ML Research Scientist
Intel Labs, Intel Corporation
02/24 – Present (1 yr 6 mo)
  • Led PM-Cobots project integrating Fanuc/UR hardware with AI perception, motion planning & autonomous workflows; achieved 50% technician load reduction.
  • Architected sim-to-real perception pipeline (Blender synthetic data + SegFormer) enabling zero-shot DIMM slot detection & autonomous insertion.
  • Developed reachability & manipulability analysis tool guiding robot model selection and precise base placement.
Senior AI/ML Research Scientist
Intel Labs, Intel Corporation
06/19 – 02/24 (4 yrs 8 mo)
  • Developed real-time human intent prediction via optimized Approximate Bayesian Computation improving task fluency & efficiency.
  • Delivered high-impact demos (Intel Innovation & Vision) showcasing low-latency 5G teleoperation, programming by demonstration & HRC.
  • Integrated ML perception (deep segmentation, probabilistic inference) with visuo-force manipulation (servoing grasping & insertion).
AI/ML Research Scientist
Intel Labs, Intel Corporation
06/16 – 06/19 (3 yrs)
  • Developed intuitive-physics system learning object dynamics from passive motion observations to anticipate trajectories & trigger actions.
  • Introduced Tree-Pyramidal Adaptive Importance Sampling (TP-AIS) and benchmarking framework for sampling-based inference.
PhD Researcher (Contact-Based Robotic Manipulation)
Universitat Jaume I
02/12 – 04/16 (4 yrs 2 mo)
  • Thesis: Contact-Based Robotic Manipulation (Cum Laude, International Mention).
  • Developed perception + reactive force/vision framework enabling robust handling of unknown objects in unstructured settings.
  • Core contributions: multimodal sensory integration; adaptive control; composable manipulation primitives; contact detection & prediction; object perception. Amazon Picking Challenge 2015 finalist.
Robotics Research Scientist
European Commission
03/08 – 02/12 (3 yrs 11 mo)
  • Advanced robotic manipulation algorithms for grasping unknown objects under uncertainty in multidisciplinary teams.
  • Projects:
    • GRASP – Cognitive grasping: planning & control for unknown objects; collaboration with neuroscientists.
    • EYESHOTS – Neuroscience-inspired heterogeneous 3D perception for manipulation.
    • GUARDIANS – Mobile robot simulation frameworks for coordinated navigation & scene detection.
Robotics Research Engineer
Spanish Science Ministry
06/07 – 12/07 (7 mo)
  • Developed algorithms for small humanoid robots.
  • Three-time winners, Spanish Humanoid Competition (CEABOT ’07, ’08, ’09).

Education

Ph.D. in Computer Science
Universitat Jaume I
2010 – 2016
  • Cum Laude (International Mention)
  • Research stays: KIT (sensor fusion ARMAR-III), Intel Labs
M.Sc. Intelligent Systems
Universitat Jaume I
2008 – 2010
  • GPA 9.13 / 10
B.S. Computer Science
Universitat Jaume I
2005 – 2007
  • GPA 8.2 / 10
B.S. Technical Eng. in Computer Systems
Universitat Jaume I
2001 – 2004
  • GPA 8.19 / 10 (Valedictorian)

Selected AI-Robotics Projects

Sensor-based Reactive Manipulation Primitives: Blind Box Emptying
Ph.D. Research
Challenge Empty a box of unknown, cluttered objects (no models / poses / count) under limited sensing.
Action Composed reactive grasp / transport / place / release primitives; compared blind random + tactile/FT exploration vs RGB-D baseline.
Result Reliable completion with only tactile/force primitives proved robustness; RGB-D added efficiency (fewer attempts, faster cycle) but not feasibility.
Sim-to-Real Scalable Perception
Project Lead & Main Developer
Challenge High cost & latency in acquiring / labeling real-world perception training data.
Action Blender-based synthetic generator with probabilistic randomizers & relational anchors; trained SegFormer for motherboard component detection.
Result Zero-shot sim-to-real DIMM slot detection & insertion; validated end-to-end perception-to-action autonomy.
Programming by Demonstration via Immersive VR
Lead Robot Control Researcher
Challenge Slow, non-intuitive authoring of multi-step contact-rich cobot tasks.
Action Mixed-reality VR teleoperation; segmented demos into parameterized, perception-conditioned primitives with automatic frame extraction.
Result Reduced programming time hours→minutes; reusable primitive library enabled multi-user consistency.
Human Intent Prediction For Fluent Collaboration
Lead Researcher
Challenge Need real-time anticipation of human intent for safe, fluent cooperative tasks.
Action Optimized Approximate Bayesian Computation pipeline delivering interactive-rate intent prediction; integrated in collaborative workflows.
Result Improved task fluency & safety via proactive adaptation; validated in realistic collaborative scenarios.
Robot Manipulability Analysis
Project Lead & Solo Developer
Challenge Heuristic, slow selection of robot model & base pose for cluttered tasks.
Action PyBullet + IK + manipulability ellipsoids; visualization, automated scoring & multi-robot/base benchmarking.
Result Accelerated layout & configuration; became standard internal tool.
Advanced Sampling Algorithms (TP-AIS)
Lead Researcher
Challenge Importance sampling methods required manual tuning & lacked robustness across distributions.
Action Invented Tree-Pyramidal Adaptive Importance Sampling + benchmarking suite against SOTA.
Result Higher efficiency & accuracy without tuning; internal adoption for probabilistic estimation.
Intuitive Physics Prediction
Research Lead
Challenge Lack of human-like intuitive physics for anticipating object motion & failure.
Action Learned physical parameters from passive observations; probabilistic simulation with multi-hypothesis rollouts; new update rules benchmarked.
Result Accurate forward predictions enabling preemptive actions; faster parameter convergence & predictive fidelity vs existing methods.
Manufacturing Cobotics (PM Buddy)
AI & Robotics Tech Lead
Challenge Preventive maintenance for LAM tools required two technicians (bottleneck).
Action Led end-to-end PM Buddy cobot development: Fanuc CRX20ia / UR20 hardware + AI stack (navigation, motion planning, CV inspection, anomaly detection).
Result Achieved 50% technician load reduction. Scalable predictive maintenance model.

Technical Skills

Programming & Systems
Python C++ Linux PyTorch NumPy SciPy Pandas Docker Git / CI OpenCV SQL Matlab URScript
Design & Tools
Blender URDF/Xacro SolidWorks 3D Printing Autodesk Fusion Manim Gimp Inkscape LaTeX
Robotics Expertise
Sensor Fusion (EKF/UKF) Point Cloud Library (PCL) Open3D KDL Differential Kinematics Inverse Kinematics Manipulation Motion Planning ROS1 \& 2 Physics Simulation Drake Gazebo PyBullet Microcontrollers
Advanced Proficient Basic

Hands-On Robotics Experience

  • Manufacturing Cobotics: Fanuc CRX20ia, UR20 – Tech lead (hardware + AI software) for PM Buddy cobot.
  • Collaborative Robotics: UR5, UR5e, Franka Emika Panda – Control algorithms & perception pipelines for demos & transfers.
  • Industrial Robotics: Baxter dual-arm manipulator – Amazon Picking Challenge 2015 team member.
  • System Integration: Dual-arm Mitsubishi PA10-7C + Schunk SDH2 + Barrett Hand + custom vision head (Tombatossals humanoid torso).

Awards & Recognition

  • 2022 – IROS Best Paper Award Finalist (Cognitive Robotics)
  • 2020 – SHPE Awards Finalist (Young Researcher)
  • 2017 – Ranked 6th, euRobotics Georges Giralt PhD Award
  • 2009–2007 – Winner, Spanish Humanoid Robot Contest CEABOT (’09, ’08, ’07)
  • 2008 – City of Castellón Youth Award (R&D+i)
  • 2004 – Best Student & Academic Excellence Awards (Technical Eng. in Computer Systems)

Invited Talks

  1. Enabling Real-Time Performance for Approximate Bayesian Computation. CIMAT-INAOE Deep Learning & Data Science Workshop, Guanajuato, Sep 2019.
  2. Robust Scene Understanding via Real-Time Approximate Bayesian Computation. IWSSIP 2020 (Online), Jul 2020.

Conference and journals

1. Felip, J. and Morales, A. (2009). Robust sensor-based grasp primitive for a three-finger robot hand. In Intelligent Robots and Systems, 2009. IROS 2009. IEEE/RSJ International Conference on, pages 1811–1816 2. Laaksonen, J., Felip, J., Morales, A., and Kyrki, V. (May). Embodiment independent manipulation through action abstraction. In Robotics and Automation (ICRA), 2010 IEEE International Conference on, pages 2113–2118 3. Chinellato, E., Felip, J., Grzyb, B. J., Morales, A., and del Pobil, A. P. (2011). Hierarchical object recognition inspired by primate brain mechanisms. In Computational Intelligence for Visual Intelligence (CIVI), 2011 IEEE Workshop on, pages 1–8 4. Bohg, J., Johnson-Roberson, M., Leon, B., Felip, J., Gratal, X., Bergstrom, N., Kragic, D., and Morales, A. (2011). Mind the gap - robotic grasping under incomplete observation. In Robotics and Automation (ICRA), 2011 IEEE International Conference on, pages 686–693 5. Felip, J., Bernabe, J., and Morales, A. (2012a). Contact-based blind grasping of unknown objects. In 12th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, pages 396–401 6. Felip, J., Laaksonen, J., Morales, A., and Kyrki, V. (2012b). Manipulation primitives: A paradigm for abstraction and execution of grasping and manipulation tasks. Robotics and Autonomous Systems 7. Leon, B., Felip, J., Marti, H., and Morales, A. (2012). Simulation of robot dynamics for grasping and manipulation tasks. In 12th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, pages 291–296 8. del Pobil, A. P., Duran, A. J., Antonelli, M., Felip, J., Morales, A., Prats, M., and Chinellato, E. (2013). Integration of visuomotor learning, cognitive grasping and sensor-based physical interaction in the uji humanoid torso. Designing Intelligent Robots: Reintegrating AI, pp. 6–11 9. Bernabe, J., Felip., J., del Pobil, A. P., and Morales, A. (2013). Contact localization through robot and object motion from point clouds. In 13th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, Atlanta, GA, USA. IEEE 10. Morales, A., Prats, M., and Felip, J. (2013). Sensors and methods for the evaluation of grasping. In Carbone, G., editor, Grasping in Robotics, volume 10 of Mechanisms and Machine Science, pages 77–104. Springer London 11. Felip, J. and Morales, A. (2014). Dual arm sensor-based controller for the cap unscrewing task. In 14th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, Madrid, Spain. IEEE 12. Felip, J., Morales, A., and Asfour, T. (2014). Multi-sensor and prediction fusion for contact detection and localization. In 14th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, Madrid, Spain. IEEE 13. Felip, J., Durán, A. J., Antonelli, M., Morales, A., and del Pobil, A. P. (2015). Tombatossals: A humanoid torso for autonomous sensor-based task execution research. In 15th IEEE-RAS International Conference on Humanoid Robots, HUMANOIDS, Seoul, South Korea. IEEE 14. Felip, J. (2016). Contact driven robotic manipulation. PhD Thesis 15. Gonzalez-Aguirre, D., Felip, J., and Shahidzadeh, S. (2016). Infinite-Resolution 3D Modeling from a Single Image for Free-Form Revolution Objects. In 2016 IEEE International Symposium on Multimedia (ISM), pages 253–260. IEEE 16. Felip, J., Gonzalez-Aguirre, D., and Tickoo, O. (2017). Towards Intuitive Rigid-body Physics Through Parameter Search. IEEE-RAS 17th International Conference on Humanoid Robotics (Humanoids), pages 0–7 17. Felip, J., Ahuja, N. A., Gómez-Gutiérrez, D., Tickoo, O., and Mansinghka, V. (2019). Real-time approximate bayesian computation for scene understanding. NeurIPS 2019. Workshop in Perception as generative reasoning: structure, causality, probability 18. Felip, J., Ahuja, N., and Tickoo, O. (2020). Tree pyramidal adaptive importance sampling. https://arxiv.org/abs/1912.08434 19. Felip, J., Gonzalez-Aguirre, D., and Nachman, L. (2022). Intuitive & efficient human-robot collaboration via real-time approximate bayesian inference. In Intelligent Robots and Systems, 2022. IROS 2022. IEEE/RSJ International Conference on. Best paper award finalist

Patents

1. Turek, J. S. et al. (U.S. Patent 20190317739 A1, 2019). Methods and apparatus to automatically generate code for graphical user interfaces 2. Leon, J. F. et al. (U.S. Patent 20190317455 A1, 2019). Methods and apparatus to generate acceptability criteria for autonomous systems plans 3. Alvarez, I. J. et al. (U.S. Patent 20190318244 A1, 2019). Methods and apparatus to provide machine programmed creative support to a user 4. Gonzalez Aguirre, D. I. et al. (U.S. Patent 20190314984 A1, 2019). Automatic robot perception programming by imitation learning 5. Remis, L. C. M. et al. (U.S. Patent 20190317965 A1, 2019). Methods and apparatus to facilitate generation of database queries 6. Felip Leon, J. et al. (U.S. Patent 20200149898 A1, 2020b). System and method for acceleration-based vector field maps 7. Gonzalez Aguirre, D. I. et al. (U.S. Patent 20200143551 A1, 2020c). Technologies for thermal enhanced semantic segmentation of two-dimensional images 8. Gomez Gutierrez, D. et al. (U.S. Patent 20200133281 A1, 2020). Safety system for a vehicle 9. Elli, M. S. et al. (U.S. Patent 20200130570 A1, 2020). Self-adaptive multiresolution digital-plate for multimodal enhanced safety of vehicles 10. Turek, J. et al. (U.S. Patent 20200130711 A1, 2020). Autonomous vehicle system for detecting safety driving model compliance status of another vehicle, and planning accordingly 11. Gutierrez, D. G. et al. (U.S. Patent 20200023842 A1, 2020). Potential collision warning system based on road user intent prediction 12. Perez-Ramirez, J. et al. (U.S. Patent 20200018814 A1, 2020b). Locating radio transmission source by scene reconstruction 13. Alvarez, I. et al. (U.S. Patent 20200326696 A1, 2020b). Technology to handle ambiguity in automated control systems 14. Felip Leon, J. et al. (U.S. Patent 20200327415 A1, 2020a). Neural network verification based on cognitive trajectories 15. Perez-Ramirez, J. et al. (U.S. Patent 20200329386 A1, 2020a). Automated network control systems that adapt network configurations based on the local network environment 16. Zamora Esquivel, J. et al. (U.S. Patent 20200327731 A1, 2020). Artificial imagination engine using 5d information descriptors 17. Gonzalez Aguirre, D. I. et al. (U.S. Patent 20200324409 A1, 2020b). Affordance-aware, multi-resolution, free-form object manipulation planning 18. Oliver, D. et al. (U.S. Patent 20200327802 A1, 2020). Object tracking technology based on cognitive representation of a location in space 19. Guzman, M. et al. (U.S. Patent 20200322528 A1, 2020). Camera orchestration technology to improve the automated identification of individuals 20. Ahuja, N. et al. (U.S. Patent 20200226430 A1, 2020). Devices and methods for accurately identifying objects in a vehicle’s environment 21. Gonzalez Aguirre, D. et al. (U.S. Patent 20200223443 A1, 2020a). Safety system for a vehicle 22. Alvarez, I. et al. (U.S. Patent 20200200869 A1, 2020a). Compensating for a sensor deficiency in a heterogeneous sensor array 23. Alvarez, I. J. et al. (U.S. Patent 20210001884 A1, 2021d). Technology to generalize safe driving experiences for automated vehicle behavior prediction 24. Alvarez, I. J. et al. (U.S. Patent 20210403031 A1, 2021a). Disambiguation of vehicle navigation actions 25. Alvarez, I. J. et al. (U.S. Patent 20210402898 A1, 2021b). Methods and devices for a vehicle 26. Felip Leon, J. et al. (U.S. Patent 20210309264 A1, 2021a). Human-robot collaboration 27. Felip Leon, J. et al. (U.S. Patent 20210299866 A1, 2021b). Robotic manipulation planning based on probabilistic elastoplastic deformation material point method 28. Datta, P. et al. (U.S. Patent 20210279506 A1, 2021). Systems, methods, and devices for head pose determination 29. Turek, J. S. et al. (U.S. Patent 20210150323 A1, 2021). Methods and apparatus to implement a neural network 30. Guim Bernat, F. et al. (U.S. Patent 20210116261 A1, 2021). Systems and methods for vehicle-occupancy-based and user-preference-based smart routing and autonomous volumetric-occupancy measurement 31. Gonzalez Aguirre, D. et al. (U.S. Patent 20210117649 A1, 2021). Systems and methods for privacy-preserving facemask-compliance-level measurement 32. Alvarez, I. J. et al. (U.S. Patent 20210114606 A1, 2021c). Systems and methods for intrusion detection in vehicle systems 33. Campos Macias, L. E. et al. (U.S. Patent 20210110264 A1, 2021). Methods and apparatus to facilitate efficient knowledge sharing among neural networks 34. Felip Leon, J. et al. (U.S. Patent 20220111526 A1, 2022). Localization system and method 35. Elli, M. S. et al. (U.S. Patent 20220114458 A1, 2022). Multimodal automatic mapping of sensing defects to task-specific error measurement 36. Campos Macias, L. et al. (U.S. Patent 20220009439 A1, 2022). Enhanced occupant collision safety system 37. Leon, J. F. et al. (U.S. Patent 20220274251 A1, 2022). Apparatus and methods for industrial robot code recommendation 38. Oboril, F. et al. (U.S. Patent 20220242452 A1, 2022). Vehicle occupant monitoring 39. Galeev, M. T. et al. (U.S. Patent 20220200920 A1, 2022). Dynamic adaptation of time-aware communications in time-sensitive systems 40. Gomez Gutierrez, D. et al. (U.S. Patent 20240217103 A1, 2024). Trajectory planning systems and methods 41. Gonzalez Aguirre, D. et al. (U.S. Patent 20240112369 A1, 2024a). Human-robot interactive workspace 42. Gonzalez Aguirre, D. et al. (U.S. Patent 20240025042 A1, 2024b). Neuro-capability plug-ins for robot task planning 43. Macias, L. C. et al. (U.S. Patent 20230259665 A1, 2023). Occupancy mapping based on geometric entities with hierarchical relationships 44. Campos Macias, L. E. et al. (U.S. Patent 20230048578 A1, 2023). Real-time predictor of human movement in shared workspaces 45. Aguirre, D. G. et al. (U.S. Patent 20240416510 A1, 2024). Object-agnostic fast grasping-points estimation via geometric-algebra 46. Felip Leon, J. et al. (U.S. Patent 20240326254 A1, 2024). Camera and end-effector planning for visual servoing 47. Gonzalez Aguirre, D. I. et al. (U.S. Patent 20240314290 A1, 2024c). Methods and apparatus to model volumetric representations 48. Turek, J. et al. (U.S. Patent 20240293931 A1, 2024). Human-collaborative robot ergonomic interaction system 49. De La Guardia Gonzalez, R. et al. (U.S. Patent 20250060729 A1, 2025). Hierarchical polytope workspace mapping for human-robot collaboration 50. Felip Leon, J. et al. (U.S. Patent 20250001605 A1, 2025b). Probabilistic autocomplete system for robot programming 51. Gonzalez Aguirre, D. et al. (U.S. Patent 20250004820 A1, 2025a). Edge-based end device control using asynchronous adaptive motion primitives 52. Turek, J. S. et al. (U.S. Patent 20250001597 A1, 2025b). Collaborative human-robot error correction and facilitation 53. Felix Rendon, J. et al. (U.S. Patent 20250128560 A1, 2025). Transport system with self-lifting wheel units for floor obstacle traversal 54. Zamora Esquivel, J. et al. (U.S. Patent 20250121493 A1, 2025). Differential kinematics control using conformal geometric entity modeling 55. Gonzalez Aguirre, D. et al. (U.S. Patent 20250135663 A1, 2025c). Illumination control in robotic end effector manipulation 56. Gonzalez Aguirre, D. et al. (U.S. Patent 20250135655 A1, 2025b). Human-robot interface system with bidirectional haptic feedback 57. Felip Leon, J. et al. (U.S. Patent 20250187181 A1, 2025a). Energy efficient robotic arm 58. Ruiz Fernandez, L. et al. (U.S. Patent 20250202788 A1, 2025). Controller parameter adaptation for non-differentiable communication conditions 59. Felip Leon, J. et al. (U.S. Patent 20250214788 A1, 2025c). Smart sensing for pallet loading and unloading 60. Turek, J. S. et al. (U.S. Patent 20250238664 A1, 2025a). Adaptation of quantization of neural network models during inference