Abstract—This paper addresses the grasp planning problem, which deals with finding the contact points between a five- fingered hand and an arbitrary object. As we consider this problem as an optimization problem, we provide in this paper an approach based on Particle Swarm Optimization for the generation and execution of grasps. Its main purpose is to compute a set of hand configurations posture in order to find an appropriate grip, satisfying a certain criteria. Assuming that the search for a solution is restricted to a precise grasp which allows only contact with the fingertips, we analyze each of the configurations of the hand with a fitness function based on a measure of quality of the grasp. Each grasp is tested and evaluated within our grasping simulator “HandGrasp”. We present experimental results using different objects. I. INTRODUCTION UTOMATIC grasp planning for robotic hands is a computationally difficult problem since it deals with the huge number of possible hand configurations. The great number of degrees of freedom of the robotic hand coupled with the space of grasping parameters and the geometry of the object creates a high dimensional space of configuration [1][2]. Besides, the dynamic modeling of the multi-fingered robot hand has to be considered [3]. The grasping process includes the knowledge of the environment. Furthermore, some information is needed whatever from the object side part like the localization, the shape, the mass, the material etc. or from the hand side part like the geometric and kinematic model, the size, etc. This information serves as input to the grasp planner. The output is the position of the fingertips on the object. In this paper, we present a fast algorithm to flexibly grasp an object using a five-fingered hand. The grasp planner is based on Particle Swarm Optimization algorithm. The main purpose of the method is to explore the dexterous manipulation space of a five-fingered robot hand and to find the best configuration of the fingers that enables a “good” grasp. This paper is organized as follows. In Section II, we present a short review of other approaches. Section III describes the geometric and kinematic model of the multi- Manuscript received January 29, 2012. The authors would like to acknowledge the financial support of this work by grants from General Direction of Scientific Research (DGRST), Tunisia, under the ARUB program. C. Walha, H. Bezine and A. M. Alimi are with REGIM: REsearch Groups on Intelligent Machines, University of Sfax, National Engineering School of Sfax (ENIS), BP 1173, Sfax, 3038, Tunisia; (phone: +216-74- 274-088; fax: +216-74-275-595; e-mail: {chiraz.walha; hala.bezine; adel.alimi}@ ieee.org). fingered hand. In Section IV, our particle swarm optimization based algorithm is presented. In order to prove the validity of the proposed approach, some simulations are illustrated in Section V. Finally, Section VI concludes the paper. II. RELATED WORK A variety of approaches [4][5] have been used to tackle the grasp planning problem notably the works of Fuentes et al. [6] who presented a grasp planner based on genetic algorithm. Also, Borst et al. [7] used a heuristic approach to plan a precision grasp for a 3D objects. Ekvall and Krajic [8] offered a simple learning and control framework for a grasping system based on human demonstration and then [9] used an algorithm based on the shape primitives of the object to generate the appropriate grasp. Pelossof et al. [10] presented an approach based on support vector machine concept involving a combination of numerical methods to recover parts of the grasp quality surface with any robotic hand in the simulator “Grasp it!”. Ciocarlie and Allen [11] presented a low-dimensional hand posture optimization method, taking into account the task, in order to generate a suitable hand posture. Li and Pollard [12] used a matching algorithm to select appropriate grasps from a database based on the shape of the object. Zhixing et al. [13] has offered a classification of grasp planners on forward and backward direction. The forward direction follows these steps:  close the fingers on the object  extract the joint angles using the kinematic model of the hand  detect the positions of the fingertips at collision, using the collision detection technique  evaluate the grasp quality. This methodology is evaluated in the simulator “Grasp it” [14][15], which have been used for analyzing and visualizing the grasps of a variety of different hands and objects. This grasp planner includes two phases, the first one is to generate a configuration of the hand using shape primitives [16], and the second one, is to evaluate the quality of these grasps. The backward direction is object centered solution and is presented as follow:  contact points are randomly or analytically located on the object surface  evaluate the grasp quality  find the corresponding feasible finger joint position using an inverse kinematic algorithm. Towards a Particle Swarm Optimization Approach for Grasp Planning Problem Chiraz Walha, Graduate Student Member, IEEE, Hala Bezine, Member, IEEE and Adel M. Alimi, Senior Member, IEEE A The Fourth IEEE RAS/EMBS International Conference on Biomedical Robotics and Biomechatronics Roma, Italy. June 24-27, 2012 978-1-4577-1198-5/12/$26.00 ©2012 IEEE 1692