: An Architecture for Knowledge Representation and Reasoning in Robotics" by Shiqi Zhang and colleagues.
In experiments with physical robots, this combined architecture reduced the time it took to finish tasks by 39% compared to traditional methods.
Using a declarative "action language" to help the robot understand qualitative knowledge and prioritized rules.
You can access the full details or download a PDF version through these platforms:
The system automatically picks out the most important details for each step, which prevents the robot from being overwhelmed by too much data. Where to Find It