Energy-aware parallel self-reconfiguration for chains microrobot networks

Highlights

• We present a parallel and energy-ware protocol for self-reconfiguration of microrobots.

• From chains configurations to squares configurations.

• This protocol is efficient and scalable because it is map-less.

• This protocol needs a constant complexity of memory usage.

• Evaluation is made with the Dynamic Physical Rendering Simulator.

Abstract

MEMS microrobots are miniaturized electro-mechanical elements, made using the techniques of micro-fabrication. They have limited energy capacity and low memory space. Self-reconfiguration is required for MEMS microrobots to complete their mission and/or to optimize their communication. In this paper, we present a self-reconfiguration protocol from a straight chain to square organization, which deals with MEMS microrobots characteristics. In the proposed protocol, nodes do not have the map of their target positions which makes the protocol portable, standalone, and the memory complexity is bounded by a constant. This paper improves a former solution by using parallelism in the movements of microrobots to optimize the time and the number of movements and by making the algorithm energy-aware. So each node is aware of the amount of energy that it will spend, which will improve the energy consumption. Our algorithm is implemented in Meld, a declarative language, and executed in a real environment simulator called DPRSim.

Introduction

Micro electro mechanical system (MEMS) is a technology that enables the batch fabrication of miniature mechanical structures, devices, and systems. MEMS are miniaturized and low-power devices that can sense and act. It is expected that these small devices, referred to as MEMS nodes, will be mass-produced, making their production cost almost negligible  [8]. Their applications will require a massive deployment of nodes, thousands or even millions  [36] which will give birth to the concept of Distributed Intelligent MEMS (DiMEMS)  [4].

The size of MEMS nodes differs from well below one micron to few millimeters. A DiMEMS device is composed of typically hundreds of MEMS nodes. Some DiMEMS devices are composed of mobile MEMS nodes  [1], some others are partially mobile  [9] whereas others are not mobile at all  [4]. Due to their small size and the batch-fabrication process, MEMS microrobots are potentially very cheap, particularly through their use in many areas in our lifetime  [10].

One of the major challenges in developing a microrobot is to achieve a precise movement to reach the destination position while using a very limited power supply. Many different solutions have been studied for example, within the Claytronics project [1], [2], [7], [25] each microrobot can only turn around its neighbor which introduce the idea of a collaborative way of moving. But, even if the power requested for moving has been lowered, it still costs a lot regarding the communication and computation requirements. Optimizing the number of movements of microrobots is therefore crucial in order to save energy  [14].

MEMS microrobots topic is gaining an increasing attention since large-scale swarms of robots will be able to perform several missions and tasks in a wide range of applications such as odor localization, firefighting, medical service, surveillance, search, rescue, and security  [8]. The self-reconfiguration for MEMS microrobots is necessary to do these tasks. In the literature, self-reconfiguration can be seen from two different points of view. On the one hand, it can be defined as a protocol, centralized or distributed, which transforms a set of nodes to reach the optimal logical topology from a physical topology  [11]. For example, if we have a connected chain of $n$ microrobots then the complexity of message exchange if a node broadcasts a message to others will be $O\left(n\right)$ in the worst case. If we reconfigure the chain to a square the complexity will be $O\left(\sqrt{n}\right)$ in the worst case. On the other hand, the self-reconfiguration is built from modules which are autonomously able to change the way they are connected, thus changing the overall shape of the logical network  [7], [31]. This process is difficult to control, because it involves the distributed coordination of a large number of identical modules connected in time-varying ways. The range of exchanged information and the amount of displacement, determine the communication and energy complexity of the distributed algorithm. When the information exchange involves close neighbors, the complexity is moderate and the resulting distributed self-reconfiguration scales gracefully with network size.

An open issue is whether distributed self-reconfiguration would result in an optimal configuration with a moderate complexity in message, execution time, number of movements and memory usage.

As said before, MEMS microrobots are low-power and low-memory capacity devices that can sense and act. A solution of self-reconfiguration should deal with MEMS microrobots characteristics. Self-reconfiguration with shared map does not scale. Because the map (predefined position of the target shape) consists of P positions and each node must have a memory capacity, at least, of P positions. Therefore, if P is very high, the self-reconfiguration will be not feasible. In this paper, we present an energy-aware parallel reconfiguration algorithm, without predefined positions of the target shape, which reduces memory usage to $O\left(1\right)$. This algorithm ensures the networks connectivity throughout all its execution time. This work takes place within the Claytronics project and aims at optimizing the logical topology of the network through rearrangement of the physical topology as we will see in the next sections.