Skip to main content
Factory of Future

M. Gulesin

Jul 1, 1993

Artificial Intelligence is a promising approach to automating process planning. Expert Systems or Intelligent Knowledge Based Systems are able to automate the reasoning activities to capture logic, experience-based reasoning and knowledge in a computer environment.

CAD/ CAM IN AUTOMATION

The developments of manufacturing can be categorized in two stages, namely the mechanization stage and the automation stage. In the early stage of manufacturing, tools and processes were mechanized. All of the various manufacturing processes were divided into categories such as casting, forging, turning, milling, drilling and cutting, with workers specialized in one of these areas. Specialization resulted in the separation of design from manufacturing. One person would design a product while other specialists would manufacture it. Design and manufacturing communicated through drawings. The mechanization stage was able to accomplish mass turnover and speed in production. However, it lacked flexibility and integration.

The next stage in the development of manufacturing is automation. In 1975, mass production was automated through the use of transfer lines. In 1976, batch production was automated through Flexible Manufacturing Systems (FMS). In 1979, design and draughting through CAD (Computer Aided Design) started to be used widely. The integration of these started in 1985. The goal of this stage is to have completely automated manufacturing plant operating with only a minimum of human involvement. Progress is being made in this regard, but total integration has not yet been achieved. The totally automated factory will be capable of mass turnover and speed in production, will be flexible and completely integrated.

The most important development regarding automation in manufacturing has been the computer. It provided developments in manufacturing control, material handling, planning and in other activities. The use of computers in manufacturing control improved NC (Numerical Control) technology such as computer aided NC code generation. It is now possible with some CAD/CAM (Computer Aided Design/Computer Aided Manufacture) systems to generate NC tape directly from the designed part stored within the CAD data base. Computers have greatly enhanced automated manufacturing. NC machine tools have been replaced by CNC (Computer Numerical Control) machine tools: almost every machining process can now be efficiently automated with a significant degree of accuracy, reliability and repeatability.

Computer Aided Design (CAD) can be defined as the use of computers to assist in the design process including calculation, analysis, modelling, draughting and testing. Initially, CAD systems were primarily used for draughting. Nowadays it also includes Finite Element Modelling (FEM), geometric modelling and kinematic analysis. (FEM is widely used for the analysis of many engineering problems, namely static, dynamic and thermal stress analysis of various structures including vibration analysis. Geometric modelling is concerned with the mathematical representation of objects in a computer.)

The development of NC machine tools was the beginning of CAM systems. CAM can be defined as the use of computers to enhance or assist in any manufacturing process. CAM comprises a large number of functions such as Computer Numerical Control (CNC), Direct Numerical Control (DNC), Flexible Manufacturing System (FMS), Automated Guided Vehicle (AGV), automated material handling, inspection and computer controlled assembly systems. CNC is an NC system that uses a dedicated computer to perform NC functions. DNC can be defined as a manufacturing system where a number of machine tools are controlled by a central computer simultaneously. The part programme is transmitted to the machine tool directly from the computer. An FMS is a programmable manufacturing system capable of producing a variety of products automatically and it is composed of CNC machine tools, automated material handling systems, robots and a computer system to control them. An Automated Guided Vehicle (AGV) is a robot-like vehicle that is used to carry objects from one place to another and can be programmed to trace a path.

INTEGRATION OF CAD/CAM

Due to development in computer technology, numerically controlled equipment, robots and computer controlled automation in CAD and CAM systems, many manual skills have been automated resulting in reduction of lead times, improvements in production, increase in manufacturing accuracy and flexibility. However, the full integration of CAD and CAM systems in industry has not yet been achieved and they have been developed separately (see Davies et al., 1988; Irani et al., 1990; Joseph and Davies 1990).

It is acknowledged that significant benefits can be obtained when CAD is integrated with CAM within a single company. For this reason the integration of CAD and CAM systems has become an important goal in factory automation (see Semakula and Gill, 1989; Sing et al., 1990). Computer Integrated Manufacturing (CIM) is the term used to denote the complete integration of all aspects of CAD and CAM systems.

CAD and CAM systems have not been totally integrated due to the difficulties in automating intermediate functions (see Joseph and Davies, 1990; Joseph et al., 1990). In order to achieve the goal of full integration of CAD and CAM two major obstacles should be addressed, namely complete CAD and CAPP/CAM interface and a fully automated, flexible CAPP system. (CAPP stands for Computer Aided Process Planning.)

CAPP is an important activity which bridges CAD to CAM (Figure 1) and translates the design information into manufacturing instructions to produce mechanical components (see, most recently Desai and Pande, 1991; Cho et al., 1991). The task of process planning in industry is usually performed by an experienced process planner manually employing his or her expertise and knowledge about machining operations. The quality of the plan developed depends on the experience and preferences of the planner whose highly skilled expertise is difficult to replace (Bandyopathyoy et al., 1981; Joseph and Davies, 1991).

Several CAPP systems have been developed. However, the computer can only assist the planner generate process plans. Fully automated CAPP does not exist and its benefits in the real industrial environment are still to be seen (see, e.g. Chang. 1990; Domazet and Manic, 1990). The complexity of decision making in process planning is a barrier to automating process planning. Many of the tasks carried out by the planner require expert knowledge, experience and intelligent reasoning (see Rustom and Mileham, 1989; Stewart et al., 1989). Other major impediments to the implementation of fully automated CAPP are related to the capturing of planning logic and heuristic knowledge. Industrial robots are programmed by a human programmer. But how does a bee know how to built a honeycomb and make honey?

 

Algorithmic programming techniques are considered unsuitable to automate process planning because process planning problems are usually solved heuristically, that is, on the basis of human ability to use reason and learn from experience (see Tonshotf et al., 1987; Dumazet, 1992). Manufacturing processes change over time on the factory floor. Algorithmic programs are not flexible enough to accommodate modifications since any alteration in the programme affects the whole structure of the software (Changer et al., 1991). Artificial Intelligence (AI) is a promising approach to automate process planning. Expert Systems or Intelligent Knowledge Based Systems are able to automating the reasoning activities to capture logic, experience-based reasoning and knowledge in a computer environment. An Expert System represents and stores the domain-specific knowledge in a special manner so that it is possible to add, delete or modify the knowledge within the database without any alteration in the program.

In short, the main goal for the industry of tomorrow is to integrate all the activities on the factory floor, i.e. to have automation from design to final manufacturing, (Nordland, 1988).

Assuming that we had a chance to visit such a factory of the future totally integrated, automated, unmanned except by robots, we would be aware that the automation is achieved and controlled by a computer programme which processes data, solves the problems that arise and gives the commands necessary to run the factory.

It is obvious that every such programme requires a programmer. Nobody would claim that the machinery, robots. etc. have themselves decided to develop the complicated software to control the system on the factory floor: it is easy to see that machines and mechanical parts do not have the ability, intelligence and knowledge even to wonder at their own structure.

Even if we do not see the programmer we can infer that one exists who is expert in the particular field and who programmed the automatic systems to do particular things. Similarly, we can liken the earth to an automated factory where animals and plants are like robots or automatic systems that perform some intelligent actions. If, within this factory a bee, for example, is not attributed to a Creator Who tells it how to make its honeycomb and honey then it must be that bees themselves know the necessary chemistry and geometry to do so. But we know that a bee is so unintelligent that when it is trapped indoors it tries to get out through a closed window. Even where there is an open window nearby it does not think of using the open window, but only finds it randomly. Therefore, we may not suppose that bees are intelligent and skilled enough to make honeycombs and honey. Even we, humans, who are the most intelligent creatures on earth, are not able to make proper hexagon-shaped honeycombs without using tools or a die. So we cannot expect a bee to do so all by itself without using a tool.

Every fruit tree is a fruit factory. A vegetable plant is a vegetable factory. They produce fruits and vegetables, respectively. If they are not attributed to a Creator then it must be that they are creating fruits and vegetables by themselves. We know that trees and vegetables are not intelligent enough and lack the knowledge of biology or chemistry to combine the necessary minerals or molecules to create the fruits and vegetables that fulfil our needs. They are not even aware of what we need. Examples can be extended to other creatures in the earth. Vegetables and animals perform some intelligent actions and yet they are not intelligent. Although we do not see the Creator of this factory, the activities around us show that there is One, Who is All-Wise, creates and controls the actions within this factory-like earth. 

REFERENCES

  • ClANG, T.C. (1990) ‘Expert Process Planning for Manufacturing’, Addison-Wesley Publishing Company, USA.
  • CHANG, T.C.,Wysk, R.A. and Wang, H.P. (1991) Computer Aided Manufacturing, Prentice Hall, USA.
  • CHO, K.K., Lee, S.H. and Ahn, J.H., (1991) ‘Development of Integrated Process Planning and Monitoring System for Turning Operation’, Annals of the CIRP, 40/1, pp.423-7.
  • DESAI. VS. and Pande, S.S., (1991) ‘GFM. An Interactive Feature Modeller for CAPP or Rotational Components’, Computer Aided Engineering Journal, pp. 217-21.
  • IRANI, R,K., Saxena, M. and Finnigan, P.M., (1990) ‘Boundary Based Feature Modelling Utility’, Proceedings of the ASME International Computers in Engineering Conference, 1, pp. 45-51, Boston.
  • JOSEPH, A.T. and Davies, B.J., (1990) ‘Knowledge Based Process Planning System for Turned Components’, The International Journal of Advanced Manufacturing Technology, 5, pp.52-65.
  • JOSEPH, A.T. and Davies, B.J., (1991) ‘Elictation of Process Planning Knowledge in a Manufacturing Environment’. The International Journal of Advanced Manufacturing Technology. 6, pp.16-34.
  • NORDLAND, G.L., (1988) ‘Integrating CAPP Into Factory Management Systems’, CAPP From Design to Production, ed. Joseph Tulkoff, SME, pp. 134-136.
  • RUSTOM, E.A. and Mileham, A.R., (1989) ‘The Development of a Generative Computer Aided Process Planning System for Prismatic Parts’, Advances in Manufacturing Technology 4. Proceedings of the 5th National Conference on Production Research, Huddersfield Polytechnic, pp. 259-63.
  • SINGH, R., Sittas, E., Mullineux. G. and Medland, A.J., (1990) ‘Intelligent Communications Between CAD and Manufacturing Activities’, Proceedings of the 28th International MATADOR Conference, pp. 305-1 2.
  • STEWART, C.D., Wallace, W. and Boswell. C., (1989) ‘The Development of a Knowledge-Based Process Planning System’, Advances in Manufacturing Technology 4, Proceedings of the 5th National Conference on Production Research, Huddersfield Polytechnic, pp. 265-68.