∗ Working now as Guest Researcher with Université Catholique de Louvain Abstract: In this paper, the basic kinematic features of a set of modular multipurpose manipulation units conceived to work as an individual or integrated system inside of a desktop assembly cell environment are analyzed and some of the specific parameters are evaluated. The basic kinematic concept of the system is presented, in which each of the components - fundamental (Mu-0) and/or auxiliary manipulation units (Mu-X) are 4dof Parallel Mechanism family members, having, symmetric and/or nonsymetric structures, but all with 2dof in-parallel actuators. The general and/or specific features of motion resulted from an analysis carried out on the identification and evaluation of kinematic parameters of fundamental manipulation unit structure and geometry. Mobility, taken into account the, number and place of the pairs, and displacements by the inverse and direct positions problems model are solved on mobile curves of guidance (MCG) equations. The work is as the first step to explore the system capabilities and consist to be a basis for further kinematic and dynamic studies. Keywords: 4dof parallel mechanisms, kinematic analysis, topology, mobility, inverse/direct positional problem, 2dof actuators, scissor, parallelograms I. INTRODUCTION The progresses in science and technology, seen as fundamental researches or product/production improved solutions, for the purpose of maintaining health, creating goods, etc or, only by increasing the every day comfort at work or home, have been always related with the development of advanced systems. In order to reach and to properly work with the constant increasing high level of values and spectrum for their prescribed (and, required) parameters, new machines, device and/or instruments - seen each or together, as a single system must include and to be conceived on innovative ideas. Their mechanical features (kinematic, dynamic, etc) have to be in detail analyzed before a practical implementation start, and correct identified (and, evaluated) as variable and constant parameters, from the performances’ level point of view. As was pointed out before, by some authors (e.g. [1], [2], etc), in the precision assembly, an (semi)automated system is required to be able to perform various simple and/or complex motions tasks over the larger work spaces with an increased performance index, including one or more, as high accuracy and speed, bigger payload capacity, etc. There are several specific tasks, which should be fulfilled by the manipulation system - transport, pick-and place, alignment, machining, soldering, gluing, sealing, inspection, etc. Excepting the standard simple operations including the transfer (transport), or simple positioning, which involve one (1) dof, most of them require more than two (2) DOF motion capability. But not always, the maximum, six (6). The manipulation system must be precise and fast enough, easy adapted at various changes related with the volume and/or type of the production/products to be fabricated. In other words, it should be flexible. These manipulation systems have to be indeed new and innovative to fulfill these requirements. They would be true (mobile) robotic systems adapted to a specific application. This will assure not only to attain, a high level of production, in automated processes, but to maintain it in (semi)automated processes helping an operator (human being) by assisting him. (And, by this to be less stressed.) A challenging tasks today, for the precision manipulation, is to conceive and design these (flexible) systems. As it was revealed from the beginning of this work [3], a gap have existed and a lack of work between micro and mini domain of manipulating small precision components at the desktop level (desktop precision devices, desktop cell, desktop factory, etc). We tried to fill it and proposed some new and/or improved solutions (e.g. [4], [5], etc). Serial stacked type based large manipulators (robots) even take in to account the last improvements regarding the direct drive solution for actuation or closed loops [6] implementations in their kinematic chain (parallelograms, for example) have proved to be still limited in performances in the terms of speed, stiffness, or accuracy, especially when we spoke about all together. In contrast, they have a good mobility and can fulfill complicated task (including large orientations) involving more the one dof. Most of the smaller size precision positioning/manipulation units (only, as example [7]) are build also on the concept of staked one axis actuated stages (X, Y, Z, goniometric, etc). The technology of producing and assembly them in standard units is well developed now, making them attractive as price. They have a small space, but good or high accuracy, depending of the design - actuation and guidance, solutions, but becomes bulky when more than 3 dof are required, including the orientation. In contrast, the X, Y, Z robotic axis (units) as for example pneumatic actuated can reach the target with high speed capabilities over large workspaces but with lower accuracy, because of intrinsic inertia. Closed loop based type mechanisms, as in-parallel actuated ones, have still enough resources and potential to substitute the serial mechanisms in accomplishing the task related with the problems from above. Parallel mechanisms (PM) as closed loop mechanisms having a mobile body (platform) in connection with fixed one (base) by a least two independent (serial) kinematics chains (legs/arms), each of them being actuated by simple actuators [8]. If the number of chains is strict equal with that of dof for the platform, PM is defined as fully parallel mechanism [9]. Applications can be seen now in almost every domain from A New 4DOF Desktop Parallel Manipulation System (D-PMS) – Kinematic Analysis and Performance Evaluation Gheorghe Olea 1* , Benoit Raucent 2 , Kiyoshi Takamasu 1 1 The University of Tokyo-Department of Precision Engineering, 7-3-1 Hongo, Bunkyo-ku, 113-8656 Tokyo-Japan 2 Université catolique de Louvain, Départment de Mécanique, Batiment Stevin, Place du Levant 2, B-1348 Louvain-la-Neuve-Belgium email: olea@ozono.pe.u-tokyo.ac.jp