Autonomous and Autonomic Swarms (original) (raw)

Requirements of an integrated formal method for intelligent swarms

2005

NASA is investigating new paradigms for future space exploration, heavily focused on the (still) emerging technologies of autonomous and autonomic systems [47, 48, 49]. Missions that rely on multiple, smaller, collaborating spacecraft, analogous to swarms in nature, are being investigated to supplement and complement traditional missions that rely on one large spacecraft [16]. The small spacecraft in such missions would each be able to operate on their own to accomplish a part of a mission, but would need to interact and exchange information with the other spacecraft to successfully execute the mission. This new systems paradigm offers several advantages: • the ability to explore environments and regions in space where traditional craft would be impractical,

Modeling NASA Swarm-Based Systems Using Formal Methods and Agent-Oriented Software Engineering

2006

NASA is researching advanced technologies for future exploration missions using intelligent swarms of robotic vehicles. One of these missions is the Autonomous Nano Technology Swarm (ANTS) mission that will explore the asteroid belt using 1,000 cooperative autonomous spacecraft. From the engineering point of view, the complexity of this kind of systems is one of the main challenges that has to be overcame, since it makes the behavior of the swarm unpredictable. In NASA, many approaches are being explored towards this goal, mainly, a tailored software engineering approach for this kind of systems, called agent-oriented software engineering, and formal methods. In this paper, we report on the main advances we have done towards modelling, implementing, and testing NASA swarms-based concept missions.

Using Formal Methods and Agent-Oriented Software Engineering for Modeling NASA Swarm-Based Systems

2007 IEEE Swarm Intelligence Symposium, 2007

NASA is researching advanced technologies for future exploration missions using intelligent swarms of robotic vehicles. One of these missions is the Autonomous Nano Technology Swarm (ANTS) mission that will explore the asteroid belt using 1,000 cooperative autonomous spacecraft. From the engineering point of view, the complexity of this kind of systems is one of the main challenges that has to be overcame, since it makes the behavior of the swarm unpredictable. In NASA, many approaches are being explored towards this goal, mainly, a tailored software engineering approach for this kind of systems, called agent-oriented software engineering, and formal methods. In this paper, we report on the main advances we have done towards modelling, implementing, and testing NASA swarms-based concept missions.

A formal approach to self-configurable swarm-based space-exploration systems

2010 NASA/ESA Conference on Adaptive Hardware and Systems, 2010

Intelligent swarms draw their inspiration from biology where many simple entities act independently, but when grouped, they appear to be highly organized. NASA is currently investigating swarm-based technologies for the development of prospective exploration missions to explore regions of space where a single large spacecraft would be impractical. The main emphasis of this research is to develop algorithms and prototyping models for self-managing swarm-based space-exploration systems. This article presents our work on formally modeling self-configuring behavior in such systems. We present a formal model for team formation based on Partially Observable Markov Decision Processes and Discrete Time Markov Chains along with formal models for planning and scheduling.

Towards a Hybrid Formal Method for Swarm-Based Exploration Missions

2005

NASA is investigating the use of swarms of robotic vehicles for future space exploration missions. Such swarms offer many advantages of traditional, single spacecraft, missions. Intelligent swarms offer potential for selfmanagement and survivability, and their emergent properties make such swarms potentially very powerful. However, they are significantly more difficult to design, and ensuring that proper behaviors will emerge is a complex task. NASA's FAST project is investigating the use of formal approaches to the specification and verification of such systems. Using ANTS, a NASA concept mission, as a case study, multiple formal methods were evaluated to determine their effectiveness in modeling and ensuring desired swarm behavior. We discuss this evaluation and propose a hybrid formal method for use in the development of future NASA intelligent swarms.

Formal Methods for Autonomic and Swarm-based Systems

Swarms of intelligent rovers and spacecraft are being considered for a number of future NASA missions. These missions will provide NASA scientist and explorers greater flexibility and the chance to gather more science than traditional single spacecraft missions. These swarms of spacecraft are intended to operate for large periods of time without contact with the Earth. To do this, they must be highly autonomous, have autonomic properties and utilize sophisticated artificial intelligence. The Autonomous Nano Technology Swarm (ANTS) mission is an example of one of the swarm type of missions NASA is considering. This mission will explore the asteroid belt using an insect colony analogy cataloguing the mass, density, morphology, and chemical composition of the asteroids, including any anomalous concentrations of specific minerals. Verifying such a system would be a huge task. This paper discusses ongoing work to develop a formal method for verifying swarm and autonomic systems.