ASCARI is a simulator dedicated to distributed and cooperative mobile robotics systems. ASCARI has been designed to be a generic framework for implementing and testing multi-agent collaborative algorithms, especially suited to evaluate algorithms performances with a non-perfect communication channel (e.g. delayed, limited bandwidth, limited range). The design process has taken into account state-of-art robotics simulators but was mainly driven by a complex new requirement: inter-agent communication has to be integrated in the simulation loop. The core of the project is a server with a simple dynamic engine, a synchronization facility and an API (Application Programming Interface) to control simulated communication which is used by server plugins to provide implementations for the characteristics of the communication channel. Different channel requirements can be added by users through the dedicated plugin. Beside the server, there are the agents involved in the simulation.The only custom code required from the user is the agent control law. The control law has to be written as a plugin which receives sensors data and control actuators using an API as an abstraction layer from hardware. As in every mature simulator, the abstraction layer quickly enables the developer to use the desired control law directly on a real robot by changing only hardware drivers. Finally, the inter-agent communication is provided in a transparent way to the user by some template communication classes. The user can exchange information between agents by simply creating senders and receivers classes with their custom data type. Simulated communications and filters are all handled automatically. The simulator is completed by a 2D viewer and a simple GUI (Graphical User Interface) that allow the user to intuitively follow the simulation evolution and to start the various processes (simulator, agents, viewer), controlling the number of agents involved in and the simulation speed. In this paper we first describe the ASCARI simulator and we then validate it on a distributed traffic control and a distributed task assignment algorithm.
ASCARI: a component based simulator for distributed mobile robot systems
FERRATI, MIRKO;SETTIMI, ALESSANDRO;PALLOTTINO, LUCIA
2014-01-01
Abstract
ASCARI is a simulator dedicated to distributed and cooperative mobile robotics systems. ASCARI has been designed to be a generic framework for implementing and testing multi-agent collaborative algorithms, especially suited to evaluate algorithms performances with a non-perfect communication channel (e.g. delayed, limited bandwidth, limited range). The design process has taken into account state-of-art robotics simulators but was mainly driven by a complex new requirement: inter-agent communication has to be integrated in the simulation loop. The core of the project is a server with a simple dynamic engine, a synchronization facility and an API (Application Programming Interface) to control simulated communication which is used by server plugins to provide implementations for the characteristics of the communication channel. Different channel requirements can be added by users through the dedicated plugin. Beside the server, there are the agents involved in the simulation.The only custom code required from the user is the agent control law. The control law has to be written as a plugin which receives sensors data and control actuators using an API as an abstraction layer from hardware. As in every mature simulator, the abstraction layer quickly enables the developer to use the desired control law directly on a real robot by changing only hardware drivers. Finally, the inter-agent communication is provided in a transparent way to the user by some template communication classes. The user can exchange information between agents by simply creating senders and receivers classes with their custom data type. Simulated communications and filters are all handled automatically. The simulator is completed by a 2D viewer and a simple GUI (Graphical User Interface) that allow the user to intuitively follow the simulation evolution and to start the various processes (simulator, agents, viewer), controlling the number of agents involved in and the simulation speed. In this paper we first describe the ASCARI simulator and we then validate it on a distributed traffic control and a distributed task assignment algorithm.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
