The Design And Implementation Of Accompanying Behavior Pattern Based Autonomous Robot Software

An autonomous robot is a goal-oriented system with various tasks that is expected to perform tasks independently in unknown environments.The software for autonomous robot usually operates upon robotic physical hardwares and computing systems,makes intelligent plan decisions for a specific user task,and drives operations of robotic sen-sors or actuators.The working environments for autonomous robots are typically open,dynamic,and non-structured which may present uncertainties and dynamics for robot plan execution.The actual results of robot plan execution may not satisfy the expected effects in planning phase,which may produce unexpected damages to the surrounding environments.Therefore,the issue of how to make robots carry out plans both robustly and adaptively to accomplish its assigned task has become an important challenge for robot plan execution in unknown environments.Meanwhile,due to complex functionalities of robot behaviors,the architecture of autonomous robot software tend to be more difficult and complex.In this case,the issue of how to effectively facilitate software design and development has become a hot issue in robotic software engineering field.To solve these issues,this paper first proposes a novel paradigm of accompanying behavior model to specify the robot behavior pattern that is able to carry out plan exe-cution process robustly.More specifically,the accompanying model generally considers the robot plan execution in dynamic environments as the result of two types of interacting behaviors with diverse aims:the task-oriented behaviors that commit to accomplish the user-assigned task,and the observation-based behaviors that dedicate to track the status of robot plan execution by continuously observing the surroundings.By abstracting two types of robot behaviors and facilitating flexible interaction schemes between these behaviors,the accompanying behavior model manages to tightly integrate the sensor inputs with plan execution process,which enables the software to keep aware of plan execution status and to adapts the plan in respond to possible run-time contingencies.Next,this paper tries to address the architectural complexity of autonomous robot software.As the accompanying behavior model introduces more frequent interactions and cooperations of two types of autonomous behaviors,the communication and interaction of software components also increase a lot,which makes it difficult and complex to program the autonomous robot software.This paper proposes to adopt the multi-agent system to model and abstract execution monitoring procedure in a single complex robotic system,which specifically focuses on the failure detection and isolation steps.In this paper,the multi-agent system views robot components as interacting agent entities and al-lows flexible and efficient coordination between robot sensors and actuators,which helps to alleviate programming efforts for software design and development.In the last,the paper proposes an autonomous robot software framework AutoRobot to facilitate engineering development of complex autonomous robot software.AutoRobot framework embeds the accompanying behavior model and multi-agent software architecture as its basic software prototype.The framework encapsulates the basic capabilities and behaviors into well-structured agent templates and interfaces,which allow programmers to use without much modifications.In the end this paper has conducted a case study that focuses on domestic services to demonstrate the feasibility and applicability of AutoRobot framework,together with aforementioned accompanying behavior models and multi-agent software architecture.