Stage
[Stage] Human-drone interaction
Date de publication
21.10.24
Prise de poste souhaitée
01.02.25
Title: Towards natural Human – Drone collaborations
The RIS team at LAAS-CNRS in Toulouse (France) is searching for several Master students interested in performing his/her internship. The internship would last between 4 to 6 months, starting in Spring 2025 and with a salary of around 600 euros.
Full internship description and list of other available internships: https://cloud.laas.fr/index.php/s/ceFfHOCt2chJGov
Project Description:
Recent progress in aerial robotics makes the application of a flying assistant closer and closer to reality, by encoding emotions into the drone behaviors [1]. Studying natural interactions between drones and humans [2] and recent works on simple interactions between a drone and a human such as simple handover procedures [3]. However, the current works are still insufficient to perform complex coordinated collaboration tasks, as it requires interaction forces between the two agents to infer the current interaction and the correct behavior to adopt. The use of aerial manipulators (drones with a mounted robotic arm) to develop the abilities of assistant drones is highly studied here at LAAS, but we are still missing key pieces for natural collaboration.
Depending on the student’s interests, different directions can be explored:
- Design and testing of a whole-body controller for an aerial manipulator, oriented on force control for interactions. The aim is to produce a stable controller compensating the arm movement on the drone, different controller type can be discussed.
- Extend our current MPC controller by incorporating a more complex system model of the drone and the Human to improve movement predictions [4].
- Extend the current works to MPPI or other optimization based controller designs.
- Explore hybrid approaches of Behavior Trees and MPC controller for online adjusting with leadership assessment based on past trajectories and behaviors [5].
Project environment:
The "Robotics and InteractionS" (RIS) is an international recognized research team at LAAS-CNRS in Toulouse, focused on developing autonomous mobile machines that integrate perception, reasoning, learning, action and reaction capabilities. The team main research areas are: Architectures for Autonomous Robots, Learning, Temporal Planning and Execution Control, and Algorithmic motion planning. RIS is composed by 8 permanent researchers, 5 PostDocs, and several Ph.D. students.
More in general, research at LAAS-CNRS spans robotics, optimization, control, telecommunications, and nano-systems. The robotics department at LAAS-CNRS counts more than 100 people and it is supposedly one of the largest and oldest robotic research department in France. LAAS-CNRS robotics department has made world-class contributions in artificial intelligence, planning, perception, humanoids, and aerial robot design and control.
Skills/Requirements:
- [Essential] Bachelor degree in robotics, engineering, applied mathematics (or related fields)
- [Essential] Solid mathematical background (ideally in robotics)
- [Essential] Scientific curiosity and “thirst for knowledge” approach for learning
- [Essential] Experience in C/C++, ROS/ROS2, Matlab/Simulink, Python
- [Essential] Very good proficiency in writing and communicating in English
- [Desired] Experience with numerical optimization tools for robotics (e.g., CaSaDi, etc.) and robotic simulators (e.g., CoppeliaSim, Gazebo, Isaac)
Useful links:
- Link to the RIS webpage: https://www.laas.fr/public/en/ris
- Link to LAAS-CNRS webpage: https://www.laas.fr/public/en
Supervision:
The project will be supervised by Dr. Marco Cognetti (GScholar, Website), and by Jonas Soueidan, a second year PhD student working on Human - Drone interactions.
How to apply:
Interested candidates are requested to apply by sending the CV and a motivation letter via email to marco.cognetti@laas.fr and jonas.soueidan@laas.fr with the subject: [Internship LAAS – HDI].
The position will remain open until satisfactory candidates are found.
Internship duration / funding / Ph.D. opportunity:
The internship will (ideally) last 6 months, and each student will receive about 600 EUR per month as stipend. The internship can lead to a Ph.D. in the RIS team if both the student and the supervisor are interested.
References:
[1] J. R. Cauchard, K. Y. Zhai, M. Spadafora, and J. A. Landay, “Emotion encoding in human-drone interaction,” in 2016 11th ACM/IEEE International Conference on Human-Robot Interaction (HRI), 2016, pp. 263–270. doi: 10.1109/HRI.2016.7451761.
[2] J. R. Cauchard, J. L. E, K. Y. Zhai, and J. A. Landay, “Drone & me: An exploration into natural human-drone interaction,” in Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing, ser. UbiComp ’15, Osaka, Japan: Association for Computing Machinery, 2015, 361–365, isbn: 9781450335744. doi: 10.1145/2750858.2805823. [Online]. Available: https://doi.org/10.1145/2750858.2805823.
[3] G. Corsini, M. Jacquet, H. Das, A. Afifi, D. Sidobre, and A. Franchi, “Nonlinear model predictive control for human-robot handover with application to the aerial case,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022, pp. 7597–7604. doi: 10.1109/IROS47612.2022.9981045.
[4] Holkar, K. & Wagh, K & Waghmare, L.. (2011). An Overview of Model Predictive Control. International Journal of Control and Automation, vol.3 no.4, pp. 47-64.
[5] L. Ruifeng, W. Jiasheng, Z. Haolong and T. Mengfan, "Research progress and Application of Behavior Tree Technology," 2019 6th International Conference on Behavioral, Economic and Socio-Cultural Computing (BESC), Beijing, China, 2019, pp. 1-4, doi: 10.1109/BESC48373.2019.8963263.