In Proc. of the 2016 Conf. on Embedded Real Time Softw. and Systems.

Robotic applications require both correct task performance and compensation for undefined behaviors. Although deep learning is a promising approach to perform complex tasks, the response to undefined behaviors that are not reflected in the training dataset remains challenging. In a human-robot collaborative task, the robot may adopt an unexpected posture due to collisions and other unexpected events. Therefore, robots should be able to recover from disturbances for completing the execution of the intended task. We propose a compensation method for undefined behaviors by switching between two controllers. Specifically, the proposed method switches between learning-based and model-based controllers depending on the internal representation of a recurrent neural network that learns task dynamics. We applied the proposed method to a pick-and-place task and evaluated the compensation for undefined behaviors. Experimental results from simulations and on a real robot demonstrate the effectiveness and high performance of the proposed method.

Any computer system must make trade-offs between freedoms retained by the system and guarantees made to the system's users. Designers attempt to balance conflicting goals, such as throughput and ease of use. Programming languages must make these trade-offs too. For example, a language with built-in garbage collection often retains the freedom to move objects around in memory, making it difficult to share objects with other processes or with hardware devices. C and C are based on an extreme set of trade-offs: In these languages, a wide variety of hard-to-avoid program behaviors, such as signed integer overflow and out-of-bounds array references, are "undefined behaviors."