Abstract – Flow time delays in Flexible Manufacturing Cells (FMCs) are caused by transport and clamping/reclamping activities. This paper shows how dynamic scheduling parameters may control the flow times of jobs and the available task windows for flow time delays. Keywords – Flexible Manufacturing Cell, Flow Time Delay, Priority Rule I. INTRODUCTION Many manufacturing firms make use of small flexible manufacturing cells (FMCs) consisting of a limited number of automated identical workstations linked by a material handling system and capable of processing different jobs simultaneously. A careful integration of such a cell in a manufacturing department may offer a company substantial competitive advantage [1]. This careful integration includes, among other things, the link of the FMC with the strategic objectives of the company, the position of the FMC in the goods flow of the manufacturing department, organizational arrangements, and the way in which the production control system of the company deals with the FMC. Important condition for a well-working production control system is the control of flow times of jobs assigned to the FMC. This paper concerns flow time control of jobs by means of selecting appropriate dynamic scheduling parameters. Many papers have been written about the scheduling and control of flexible manufacturing systems [2]. A limited number of these papers refers to real-life systems. Few papers propose scheduling approaches that are really implemented or that are close to implementation. Examples of these papers are [3] and [4]. In our study, we decided to start from a real-life situation and to build our scheduling approach on an accurate analysis of the operation of the FMC. Furthermore, we adopted a dynamic scheduling approach close to the way in which the FMC has been controlled by the operators. This approach has given us insight in some important aspects which are not yet covered in the available scheduling literature. Based upon our study, we conclude that an optimal use of an FMC requires a carefully chosen dynamic scheduling approach which considers not only machining activities, but also the activities of transport vehicles and operators. These latter activities are usually not included in scheduling approaches presented in the literature. They, however, cause serious flow time delays of jobs. Caprihan et al. [5][6][7] provide a detailed account of the manifestation and scheduling impact of related (information) delays within manufacturing systems. In this paper we will focus on the linkage between our scheduling approach and the performance of an FMC. It will be made clear that the proposed dynamic scheduling parameters have an important impact on the available task windows for flow time delays. Our industrial case concerns the FMC of a company in Hengelo, The Netherlands, which produced pneumatic actuators for valves. The company first installed an FMC consisting of two machining centers (see Fig.1). Later, the company integrated a third machine. We developed the scheduling approach for the three-machine case. The FMC of Fig. 1 consists of two basically identical machining centers linked together by a pallet transport vehicle and an integrated pallet buffer system with a capacity of 15 pallets. Each of the machines can hold a limited number of tools in a local tool store. Tool changing within the tool stores is performed automatically. The linkage between the central tool storage and the local tool stores is realized by means of a tool robot. An FMC- computer is responsible for the coordination of all activities within the FMC. The clamping and unclamping of parts on pallets/fixtures is done manually on one or more integrated clamp/unclamp stations. Before starting the manufacturing of a job on the FMC, an operator has to (i) prepare the required pallet(s)/fixture(s) for the specific job, (ii) build up and pre-set the required cutting tools, (iii) load the central tool storage with these tools, and (iv) instruct the FMC-computer as for job release. Next, one or more items of the job can be clamped on a pallet/fixture. This is done at a clamp/unclamp station (C1, C2). The pallet transport vehicle transports the palletized items either to the pallet pool (1-15) or directly to a machining centre (M1, M2). After machining, the pallet transport vehicle transports the palletized items either to the pallet pool or directly to a clamp/unclamp station. At this clamp/unclamp station the items are replaced by new items. The refilled pallet/fixture repeats its cycle through the system. The replaced items may wait outside the FMC for being clamped on a pallet/fixture for their next operation. Section II of this paper introduces some major dynamic scheduling parameters which may play a role in the scheduling of operations on the FMC. Section III illustrates, by means of a Gantt Chart, how these parameters determine the timing of operations on the FMC. This section also presents the performance indicators by which the scheduling result can be evaluated. These indicators are, as the scheduling parameters, inspired by industrial practice. Section IV describes the experimental design for investigating the impact of the various scheduling parameters. Results are presented in Section V. Controlling Flow Time Delays in Flexible Manufacturing Cells J. Slomp 1 , R. Caprihan 2 , J.A.C. Bokhorst 1 1 Faculty of Business & Economics, University of Groningen, Groningen, The Netherlands 2 Faculty of Engineering, Dayalbagh Educational Institute, Dayalbagh, Agra, India <j.slomp@rug.nl>, <rcaprihan@gmail.com>, <j.a.c.bokhorst@rug.nl> 978-1-4244-4870-8/09/$26.00 ©2009 IEEE 149