[Master Thesis] Parallel elastic actuator for efficient manipulation
02.10.2024, Diplomarbeiten, Bachelor- und Masterarbeiten
Application deadline: October 20, 2024
Task of this thesis is development of Perallel Elastic Actuator for experimentation and application of novel control framework exploiting actuator intrinsic elasticities.
Robotic solutions recently have been largely getting accustomed to the Direct-Drive actuation, involving lower gear ratios and BLDC motors. These solutions offer several advantages:
- Good back-drivability
- Joint torque sensing based on the motor currents
- Good torque density, due to the specificity of BLDC motor
An article introducing as they call “Proprioceptive” concept for actuation (Direct-Drive using BLDC) [1] (MIT Mini Cheetah-like robots have become a popular choice for a legged locomotion research system.)
However, Direct-Drive actuators have not become so popular for the Manipulator applications:
- Consuming big amount of power to compensate for gravitational loads when stationary
- Nature of manipulator control is not impact driven (as is the case for the legged robot walking), thus more traditional designs with additional torque sensors have good enough bandwidth for common manipulation tasks)
- Torque sensing is not required for adopted industry level manipulator applications
There are examples of the Manipulator robots that adopt direct drive actuators (of similar design to cheetah) https://rsl.ethz.ch/robots-media/dynaarm.html For example, to overcome overheating problems (power consumed constantly due to active gravity compensation), they adopt the active cooling methods.
Parallel elastic Actuator
Since the manipulator/or legs often require just relatively small operational range, idea is to utilize parallel spring in order to alleviate some of the load from the motor. Drawback is that for certain motion, beside overcoming the load required to move the robot, actuator needs to “fight” also spring force. However, if due to the spring choice, one can mostly assume that the main goal of the spring is full or partial compensation of the static loads, this scenario offers more advantages (we argue).
Research question
Advantage of this topic is that it has high relevance for the research (contribution): Control of the parallel elastic actuation to overcome the negative spring properties such as oscillation. Explore damping capabilities of the actuator and utilization of the spring for passive control.
What you will gain:
- Hands-on experience and in-depth understanding of Classical Robotics and various control methods
- Best practices for software development and collaboration
- Experience building, prototyping
- Insights in our System Development and access to our community
Requirements from candidates:
- Mechanical Engineering background
- Matlab skills
- Any CAD software for the Part designs (such as Solidworks, Fusion 360,etc.)
- Basic experience in C (embedded) programming
- Basic skills in Electronics
- Plus are:
- Understanding how Motors work
- Familiarity with GIT
- Working skills in Ubuntu operating system
To apply, you can send your CV, and short motivation to:
Supervisor
M.Sc. Vasilije Rakcevic
[1] P. M. Wensing, A. Wang, S. Seok, D. Otten, J. Lang and S. Kim, "Proprioceptive Actuator Design in the MIT Cheetah: Impact Mitigation and High-Bandwidth Physical Interaction for Dynamic Legged Robots," in IEEE Transactions on Robotics, vol. 33, no. 3, pp. 509-522, June 2017, doi: 10.1109/TRO.2016.2640183.
Kontakt: vasilije.rakcevic@tum.de