Autonomous skill-centric testing using deep learning

Abstract

Abstract-Software testing is an important tool to ensure software quality. This is a hard task in robotics due to dynamic environments and the expensive development and timeconsuming execution of test cases. Most testing approaches use model-based and / or simulation-based testing to overcome these problems. We propose model-free skill-centric testing in which a robot autonomously executes skills in the real world and compares it to previous experiences. The skills are selected by maximising the expected information gain on the distribution of erroneous software functions. We use deep learning to model the sensor data observed during previous successful skill executions and to detect irregularities. Sensor data is connected to function call profiles such that certain misbehaviour can be related to specific functions. We evaluate our approach in simulation and in experiments with a KUKA LWR 4+ robot by purposefully introducing bugs to the software. We demonstrate that these bugs can be detected with high accuracy and without the need for the implementation of specific tests or task-specific models.

A novel Skill-based Programming Paradigm based on Autonomous Playing and Skill-centric Testing

“…Our contribution is an integrated platform for robot development based on technology for autonomous robotics. We combine previous work on autonomous skill acquisition and robot software testing [1]- [3] in a unified framework. We further extend our framework by a visual programming capability in which a basic behaviours can be implemented in a simple way.…”

“…In this section we sketch our skill acquisition paradigm based on technology for developmental robotics. We provide a formal definition of behaviours and skills by unifying problem formulations from previous work [1]- [3], c.f. section IV-A, and give a high-level description of our framework, c.f.…”

“…For self-diagnosis module we use a method for autonomous testing developed in previous work [3]. During a skill execution, an observation o = (M σ (s), F σ (s)) is made.…”

“…This range is extended by autonomous playing, where sequences of behaviours are used to prepare the environment such that the basic behaviour can be applied successfully. We further introduced a novel approach for autonomous skill-centric testing which enables even non-experts to automatically run tests on robot software without the need for designing unit tests [3].…”

“…We propose a novel programming paradigm based on our previous work on autonomous robots [1]- [3]. We extend our system to support visual programming.…”