Kittl, B. & Baumann, C.
Abstract: Due to changes in the manufacturing environment, concepts and EDP components currently used for shop floor data acquisition no longer meet different requirements. This paper is to set out new demands on shop floor data acquisition systems, with a special emphasis not only on the interface between the planning level and the workshop, but also on horizontal integration of applications implicated in job processing.
Keywords: Shop floor data acquisition, production control
1. INTRODUCTION
Production control and monitoring must be based on up-to-the-point and exact information from the shop floor (Mendoca et al., 1992). This includes status information on manufacturing jobs, assembly and testing jobs, as well as data on job processing resources, such as machinery, manufacturing aids, personnel and material. However, a dynamic environment also implies that, when deviations from the set plan necessitate a change in job specifications, the altered data should be supplied to the workshop immediately and distributed without delay, thus forming a closed control loop. In addition to vertical integration of the control level and workshop, modern systems must ensure horizontal integration of all EDP applications and computerized manufacturing installations involved in job processing. Response to events at manufacturing level can consequently be sped up by shortening control loops, thus reducing turnaround time. Conventional shop floor data acquisition systems fail to fulfill these requirements and need to be substituted by open shop floor communication systems.
2. PROBLEM STATEMENT
Shop floor data acquisition is nowadays seen as the collection of feedback to jobs currently processed in the workshop and the acquisition of parameters on production resources (machines, personnel etc.). This concept and the systems presently in use for shop floor data acquisition have the following shortcomings:
Unidirectional communication: Although most enterprises nowadays attach great importance to online capture of feedback from the shop floor. Target specifications (operations released) and related information are generally still provided and issued manually in the form of data sheets. This leads to a breach in the closed control loop intended to link the planning and control level with the workshop. Manual data distribution leads to delays in the availability of new target specifications at the shop floor. Feedback information registered by the shop floor data acquisition system cannot be used to update and display job process information for the workshop personnel. Job process status is thus still read from the job ticket only. Essential information supplied by shop-related applications (e.g. availability or provision schedules for tools required for a given job; availability of base material) cannot be accessed from the operators workplace.
Application-specific data acquisition: Communication between CAM functions involved in job processing is confined to direct linkage of individual planning or control functions by means of specially designed data capture devices (job processing time terminals report to PPC, personnel time & attendance terminals supply data to wage accounting, while devices capturing machine data communicate with maintenance applications). Significant information must be gathered retrospectively as there is no automatic linkage of area-specific feedback data, (e.g. job-related machine or tool faults, relations between process parameters and quality data).
Defective data: Due to the simple conception that marks conventional shop floor data acquisition systems, plausibility checking of feedback data in these systems is insufficient, if provided for at all. Often there are delays between input and processing, particularly if the data has been captured manually. Consequently, only formal criteria can be applied to check feedback data. The data acquired must be checked and corrected retroactively, which consumes additional time and resources.
Insufficient communication between shop floor data acquisition systems and shop floor EDP applications: Most shop floor data acquisition systems do not allow for interfacing with other applications or automatic manufacturing devices. Dialog processing systems are used for planning and control, whereas computer-controlled machinery is employed in manufacture. Communication between the two levels is still implemented manually, i.e. by workshop personnel gathering data from one subsystem to subsequently enter it into another one.
3. MANUFACTURING ENVIRONMENT
Conventional shop floor data acquisition systems fail to fulfill requirements resulting from changes in the manufacturing environment.
Turnaround time: For most production enterprises, short delivery periods have become a key factor for success, which has led to increased pressure on job turnaround times. Wherever possible, fault conditions and deviations are to be compensated directly at the shop floor. Essentially, this requirement implies that complete and up-to-date information on jobs currently processed must be available, and that inter-application data communication is to be ensured (job process monitoring, quality checking, tool provision etc.).
Flexibility in setting job specifications: Jobs often need to be altered, cancelled, suspended or rescheduled even after release to the workshop. This, in turn, requires distributed EDP systems that swiftly distribute and supply current order data in order to assure immediate availability of essential information directly at the workplace, or at the organizational units responsible for providing tools and material.
Automatic processing: Automatic manufacturing cells allow for quasi-parallel processing of different workpieces for several jobs. In this case, automatically gathering data from the cell control system constitutes the only feasible way to register job-assigned set up and processing time.
Efficient production organization: Increased pressure to reduce turnaround time and cut non-productive activity is giving rise to lean organization structures. Responsibilities are increasingly transferred to the workshop floor personnel; operators may decide on short-term control measures and are put in charge of organizing the workplace or manufacturing cell. However, any transfer of decision-making power and responsibility presupposes for the requisite data and information to be available at the workplace.
4. DEMANDS ON SHOP FLOOR DATA ACQUISTION SYSTEMS
A system that meets the requirements set out above is to manage all the job information and resources data to be taken into account in job processing; it will dynamically distribute this information to adjacent applications as well as user and machine front ends, link the information to feedback information online, and thus generate effective shop floor data for planning and scheduling. Demands on modern shop floor data acquisition systems may thus be defined as follows:
Data integration: Data objects taken over from planning level must be specified in the workshop (e.g. selecting a workplace from the capacity node, entering commencement and completion times for consecutive operations) and related to other objects not taken into account at planning level (allocation of tool provision or transport jobs to operations released; status of machinery and installation). This implies creating a shop floor data base capable of managing and logically linking all objects in manufacture required for job control (Jin et al., 1999). Since these objects are generated and processed by different applications, the different structures, views and degrees of specification need to be unified in the shop floor data base. Once scheduling data has been supplemented and specified according to shop floor requirements, it is immediately linked to feedback data and checked for plausibility. The resulting data constitutes the shop floor data as such, which can be viewed form all workplaces and transferred to the planning level.
Dynamic information distribution: Unlike conventional client-server structures with data being queried from a central server, shop floor communication systems call for dynamic structures. Status changes of an object managed by the shop floor data acquisition system must be communicated to all system front ends affected. For instance, operation data displayed at the workplace terminal must be automatically updated each time a given precursory operation is reported completed, or as soon as the tools requested for the operation have been supplied at the tool store. This ensures that the workshop personnel has access to all the necessary information on the operations due to be processed. At the workplace terminal, the workshop operator is updated on jobs or operations at hand, on availability of base parts or materials, and on secondary resource capacities. On the other hand, the terminal allows for the operator to generate requirements in turn (tool provision, NC program download) and enter feedback information on job process status.
Inter-application communication: The shop floor receives information from several production planning and control applications, from shop floor applications such as transport and tool management systems and from machine data acquisition systems. Due to close inter-object linking, conventional inter-application file interfaces are unsuitable. Consequently, the shop floor communication system is to provide adequate functions to the applications involved in job processing, allowing for individual objects to be generated, altered or deleted. A PPC system can thus release jobs or operations for the workshop, which are subsequently supplemented by scheduling data from a control panel. Tool provision jobs are then generated in the workshop and assigned to the individual operations. Finally, feedback from the different subsystems is gathered and linked to the order data. Apart from providing the necessary interfaces, the shop floor data acquisition system must also fulfill the function of managing applications access rights governed by factors of time and logic (who is allowed to generate new objects, who may alter them in given circumstances).
Workflow functions: Likewise, new requirements necessitate the integration of workflow functions that automatically react to changes in the status of objects managed by prompting adequate activities (e.g. by generating a transport job once an operation is reported accomplished, and releasing the subsequent operation upon receiving transport feedback).
Free configuration of user front ends: User front ends must be freely configurable to adapt to the individual workshops organizational structure and different workplace types. A number of parameters must be set: defining the users of a terminal and their rights (planning/scheduling functions or feedback only); assigning the terminal to a given manufacturing area, capacity node or workplace; fixing the maximum number of users and operations that may be registered simultaneously, allocating machinery and installations that automatically generate status messages; defining the type of feedback, and ensuring central configuration of all security and access rules on both system and user interface levels.
5. STATE OF DEVELOPMENT
Following the definition of requirements to be met by shop floor data acquisition systems geared to the present manufacturing environment, a prototype has been developed and implemented at the Institute for Production Engineering (Baumann & Kittl, 1999). This prototype is based on the following technologies:
6. REFERENCES
Baumann, Ch., Kittl, B. (1999). New Software- and Networking-Concepts bring more Flexibility to Shopfloor Data Acquisition Systems, Proceedings of 10th International DAAAM Symposium, Katalinic, B. (Ed.), Vienna, October 1999
Jin, J.Q. et al (1998). A method for the acquisition of users requirements in discrete manufacturing cell systems. Computer Integrated Manufacturing Systems, Vol. 11, No. 3, p.229-242, 1998
Mendonca, J.M. et al (1992). Integration of Shop Floor Control Applications in a Real-Time Data Collection Environment, Proceedings of Eighth CIM-Europe Annual Conference, Birmingham, U.K., 1992
Authors:
Dr. Burkhard Kittl, Dr. Christian Baumann, E-mail: kittl@mail.ift.tuwien.ac.at,
baumann@mail.ift.tuwien.ac.at
both: Institute for Production Engineering, Vienna University of Technology,
Karlsplatz 13 / 311, A-1040 Wien