The swarm intelligence paradigm has proven to have very interesting properties such as robustness, flexibility and ability to solve complex problems exploiting parallelism and self-organization. Several robotics implementations of this paradigm confirm that these properties can be exploited for the control of a population of physically independent mobile robots. The work

While many methods to simulate flocking behaviors have been proposed, these techniques usually only provide simplistic navigation and planning capabilities because each flock member's behavior depends only on its local environment. In this work, we investigate how the addition of global information in the form of a roadmap of the environment enables more sophisticated flocking behaviors and supports global navigation and planning.

Flocking behavior is very common in nature, and there have been ongoing research efforts to simulate such behavior in computer animations and robotics applications. Generally, such work considers behaviors that can be determined independently by each flock member solely by observing its local environment, e.g., the speed and direction of its neighboring flock members. Since flock members are not assumed to have global information about the environment, only very simple navigation and planning techniques have been considered for such flocks.

Abstract. A swarm-bot is a robotic entity built of several autonomous mobile robots (called s-bots) physically connected together. This form of collective robotics exploits robot interactions both at the behavioral and physical levels. The goal of this paper is to analyze the physical performance of a swarm-bot as function of its size (number n of s-bots composing it). We present three tasks and the corresponding swarm-bot performances. In all three tasks we show superlinear performances in a range of n where the physical forces applied in the structure fit to the robot design. This superlinear performance range helps in understanding which swarm-bot size is optimal for a given task and gives interesting hints for the design of new application-oriented swarm-bots. 1

Abstract – This paper presents the latest developments around our mobile micro-robot Alice. This small robot is the starting point driving and enabling enhancements in locomotion, energy, communication, perception and control. A set of new features and new HW modules is described. The robot itself is ~2x2x2 cm 3 and is able to move, sense, receive remote commands and locally communicate with neighbor robots. Extension modules implement a long-range sensor, radio communication, a linear camera and an energy pack. The current projects in biomimetic, collective and evolutionary robotics using this set of tools are also shortly explained. Index Terms – Mobile robotics, micro robotics, modular systems, micro-mechatronics.

We study the cooperative transport of a heavy object by a group of mobile robots. The system is fully decentralized and the information flow between the robots is limited to physical interactions. The robots have no explicit or implicit knowledge about their relative positions.

While techniques exist for simulating group behaviors, these methods usually only provide simplistic navigation and planning capabilities. In this work, we explore the benefits of integrating roadmap-based path planning methods with flocking techniques. We show how group behaviors such as exploring can be facilitated by using dynamic roadmaps (e.g., modifying edge weights) as an implicit means of communication between flock members. Extending ideas from cognitive modeling, we embed behavior rules in individual flock members and in the roadmap. These behavior rules enable the flock members to modify their actions based on their current location and state. We propose new techniques for three distinct group behaviors: homing, exploring (covering and goal searching) and passing through narrow areas. Animations of these behaviors can be viewed at

Abstract. While techniques exist for simulating swarming behaviors, these methods usually provide only simplistic navigation and planning capabilities. In this review, we explore the benefits of integrating roadmapbased path planning methods with flocking techniques to achieve different behaviors. We show how group behaviors such as exploring can be facilitated by using dynamic roadmaps (e.g., modifying edge weights) as an implicit means of communication between flock members. Extending ideas from cognitive modeling, we embed behavior rules in individual flock members and in the roadmap. These behavior rules enable the flock members to modify their actions based on their current location and state. We propose new techniques for several distinct group behaviors: homing, exploring (covering and goal searching), passing through narrow areas and shepherding. We present results that show that our methods provide significant improvement over methods that utilize purely local knowledge and moreover, that we achieve performance approaching that which could be obtained by an ideal method that has complete global knowledge. Animations of these behaviors can be viewed on our webpages. 1

Abstract. Scaling down robots to miniature size introduces many new challenges including memory and program size limitations, low processor performance and low power autonomy. In this paper we describe the concept and implementation of learning of a safewandering task with the autonomous micro-robots, Alice. We propose a simplified reinforcement learning algorithm based on one-step Q-learning that is optimized in speed and memory consumption. This algorithm uses only integer-based sum operators and avoids floatingpoint and multiplication operators. Finally, quality of learning is compared to a floating-point based algorithm. Keywords: Reinforcement Learning; Q-Learning; micro-robots. 1.

Scaling down robots to miniature size introduces many new challenges including memory and program size limitations, low processor performance and low power autonomy. In this paper we describe the concept and implementation of learning of a safe-wandering task with the autonomous micro-robots, Alice. We propose a simplified reinforcement learning algorithm based on one-step Q-learning that is optimized in speed and memory consumption. This algorithm uses only integer-based sum operators and avoids floating-point and multiplication operators. Finally, quality of learning is compared to a floating-point based algorithm.