|
Briefing
|
- Machine Learning Method – Australian and Chinese researchers authored paper describing new method that uses machine learning and algorithms to enable humanoid robot to find its balance in moving platform
- Hybrid Learning Models – Uses both model-based learning which gives robot different models it can use to behave in certain scenarios and model-free learning
- Successful Simulation – Shows algorithm was able to stabilize robot in platform under unknown rotations
- Implications – Demonstrate robots’ ability to adapt and walk in changing environments and unpredictable terrains outside of labs, useful in healthcare, education, disaster response, and entertainment
- Future Plans – Include testing method in more complex environments as well as enabling robots to imitate human motion, like ankle joint movement, without feeding prior information or input
|
|
Accelerator
|
|
|
Sector
|
Information Technology
|
|
Function
|
Research and Development
|
|
Organization
|
Monash University, The University of Sydney
|
|
Source
|
|
|
Original Publication Date
|
August 23, 2019
|
