Automation: Goods

Section III

Current Advances in Manufacturing Automation

Diversification created problems that have occupied manufacturing innovators for over 50 years. Producing multiple products in a single factory creates at least two major problems to be overcome:

1. Getting the right input to the right place at the right time.

2. Determining the best production run to match supply with demand.

The input can be a part or work in progress for discrete production or a combination of chemicals in continuous production. The simplest solution to this coordination problem is to stockpile inventory of inputs at each station in the production process to ease the timing and quality control problems. Assemblers can find a correct part in the inventory. This simple solution has a hidden cost in that the inventory or parts and work in progress is a financial investment that garners no rate of return until the firm obtains the payment for the sale of the final product.

The economics of the second problem are determined by the cost and time to changeover from the production of one product to another. The more costly the changeover and the longer the required time, the longer the production run to distribute the fixed costs of changeover. But again, long production runs also have the hidden cost in that the inventory of final products does not garner a rate of return until payment from the sale. Finally, the faster and cheaper the changeover, the greater the variety of products that can be produced at a single factory.

Let us start by considering the status of automation in manufacturing. Manufactured items are continuous, such as liquids; or discrete, such as automobiles. The current status of production in manufacturing is:

a. Continuous process fluids: Chemicals, beer, petrochemicals. These types of production, whether batch or continuous, are currently highly automated. So called production workers sit around and watch the dials.

b. Discrete: In discrete production the size of the production run determines the efficiency of the process.

(1) Mass: In high volume production it pays to have a specific machine for each purpose. The precise number that constitutes high volume depends on the type of product. Automobiles are a good example because production runs are generally in excess of 200,000 units.

(2) Batch processing: In batch production the production lots run from 10 to 1000. Examples of batch production are airplanes, large earth moving equipment, and ships. Batch processes comprise 40%of the mfg work force. In batch production general purpose machines are used instead of the specific purpose machines of mass production. The cost of batch production is 10- 30 times the cost of equivalent mass production.

(3) Individual production: This type of production exists today only for artisan items. The cost is 100 times as much as mass production. For example, compare the cost of auto repairs with the cost of the original production. How much damage does it take to total a car?

Steps in discrete manufacturing

To discuss automation of discrete manufacturing we need to breakdown production into its components:

a. Design: Buzz words - CAD, computer assisted design; CAE, computer assisted engineering

b. Parts manufacture: The buzz word here is flexible manufacturing systems, FMS, which are also called manufacturing cells. It is very important for the student to realize that a FMS is really a computer-controlled machine shop and is not a complete automated manufacturing plant.

c. Parts coordination: To assemble a durable good you must get the right part to the right place at the right time. When you consider that auto production involves tens of thousands of parts, this is no easy matter.

d. Quality control: To reduce waste, inventory of parts and work-in-progress, it is necessary to greatly improve quality in all aspects of the manufacturing process. The result is a high quality product.

e. Assembly: Here we are talking about the assembly line. To automate assembly requires much more than replacing people with robots. Efficient use of robots usually requires a complete reorganization of production.

f. Integration: Buzzword - CIM, computer integrated manufacturing. Currently automation is proceeding piecemeal in each area. Advances in computation and communications provide the building blocks. Integration of the steps into fully automated production will take time. The various steps have incompatible standards so communication is difficult.

g. Reorganization: Innovation in manufacturing is much more than simply substituting machines and software for humans in the production process. As automation advances, firms must also constantly innovate by reorganizing their human-machine production process to achieve an edge in international competition.

Now let us consider each step of discrete manufacturing automation in detail.

Design and Analysis

In order that the reader relate the various concepts being developed, we want to discuss the general trends in CAD/CAE with respect to Moore's law and the development trends in software in general.

General trends in CAD/CAE software

1. FROM MAINFRAME TO PC: The first CAD program was a mechanical drafting program created by GM in the 60s. It cost $1.2M and required a mainframe. As computer power has constantly increased due to Moore's law, CAD/CAE software has gravitated to PCs with a corresponding decrease in cost.
2. CONSTANTLY IMPROVING GRAPHICS: The improvement in graphics in CAD/CAE parallels the general improvement in computer graphics as computers have become more and more powerful.
3. Rather than constantly reinvent the wheel, design software is increasingly developing libraries of basic components and operations relevant to the industry.
4. SOFTWARE INTEGRATION: As business software has progressed from individual software programs to office suites so has CAD/CAE software progressed from individual programs to design suites that integrate:
   1. All aspects of the design process
   2. CAE with CAD
   3. CAD/CAE with manufacturing
5. COLLABORATIVE SOFTWARE: Lotus Notes was the first software to foster collaborative efforts of business people, engineers are also developing collaborative software to promote collaboration among the design team.
6. INCREASED SPECIALIZATION: As the market for CAD/CAE software has grown standards are constantly being set and revised. The major players focus on the largest design markets and the niche players focus on provided specialized software for niche markets. Software for CAD/CAE has become much more specialized than just having specialized software for the various engineering disciplines such as architecture, aeronautical, civil, electrical, mechanical, and mining. An engineer designing the electrical circuit for an IC would use different software than an engineer designing the electric system for an automobile.

Specific Industries: Now let us consider CAD/CAE from the perspective of specific industries. This discussion focuses on examples and definitely does not cover the vast array of software for CAD/CAE.

• Architecture
  • List of popular architecture CAD programs
• Aeronautical
  • Boeing designed its 777 as the first 100% digital airplane. The design of the 787 Dreamliner in worldwide
  • One of the many CAE tasks in designing an aircraft is determining the lift and drag characteristics of the wing. Hanley Innovations provides software to analyze model airplane airfoils on a PC. Notice how accurate the simulations are in comparison to wind tunnel data. Check out NASA's Virtual Windtunnel to see how visualization increases the power of a CAE tool.
  • In the 60s, your Prof was an aerospace engineer and developed software to optimize the payload of the Saturn SIVB into orbit before lunar launch.
• Civil
  • List of civil engineering software
• Electrical
  • Electrical system design for equipment such as automobiles and buildings. Check out the Scada electrical design system.
  • ICs
    • First realize that there are numerous types of ICs such as pure digital, pure analog, mixed digital and analog based on more than one type of transistor technology such as CMOS and Bipolar. There are also optical ICs and in the very near future there will be an explosion of mixed mechanical (nanotech) and electrical ICs. The market for software to design ICs is over $3B a year.
    • A simplified view of the design process for an IC is:
      • Create a library of basic components. For example, an adder would be a basic component for a digital design software. This design process starts with Boolean algebra. The industry is moving to intellectual property on basic components that have been optimized. For a discussion of a library see ARM . Think about the economics as avoid reinventing the wheel. It is cheaper to buy or license a good design than to reconstruct it from scratch avoiding intellectural property rights.
      • Design the circuit in terms of modules and then simulate the behavior of the circuit to test the design for flaws, for example see Mentor Graphics products. To simulate a design using HDL the designer might use a Synopsys product.
      • Compile the design into actual electronic components in silicon. Look as Cadence's software products for a discussion of this phase of IC design. There are tradeoffs in terms of such factors as minimum space, signal integrity, and heat buildup. The circuit is simulated to check its performance before actually constructing a prototype. Every software tool that allows you to remove problems accelerates the design to market process. There are many software tools to simulate the performance of an IC and analyze for potential problems. See Synopsys's verification programs as an example.
  • Mechanical
    • An important CAE tool in mechanical and aeronautical engineering is finite element analysis that computes stress on physical objects. For a list of FEA software click here. In the development of collaboration tools an example is virtual product development
    • Let us consider the design of automobiles, an industry with perhaps the most highly developed design and simulation software.
      • Check out car design online (current)
      • IBM's Product Lifecyle Management. Latest success story. Holographs are now used in car design.

Innovation: Now let us consider why the software developments in CAD/CAE constitute an innovation.

• Feasibility: Some design processes simply would not be feasible without software/computer assistance. For example, suppose you were going to lay out an IC with 1B (US billion) electronic components by hand and it only took you 1 second to draw the component on paper by hand. It would take you 2,777,778 hours or 317 years to complete the task. Yes, you could have a 1000 person design team, but think about the coordination problems to produce an error free design.
• Prevent Murphy's Law: All departments in the firm will have the most current design. Before computer stored designs, each department might have a different version creating opportunities for Murphy's law to occur.
• Design families of products efficiently: The engineer can start with the existing designs and modify them. Because a substantial percent of the new parts can be obtained through modification of the old parts, great savings of time and engineering effort are obtained. Chrysler's consolidation of their K-car product line reduced the number of parts from 75 to 40 thousand. The use of computers allows for the clustering of similar parts in design and production. Only 20% of new parts require new designs and another 40% can be created by modifications of old designs. This saving is achieved by better organization through computer use.
• Speed up design to market process: Part of the speedup in the design to market process is in better organization in granting the design team leader the authority to make decisions and have members for various departments in the firm ensure that the design can be manufactured and sold. But, an important aspect of the speedup is improvement of software tools. It is much faster and cheaper to simulate a process in a computer than to make a physical prototype and perform actual tests on the prototype. One example of this in the paperless design of automobiles and airplanes is that it is possible to see how well the parts will fit together in a computer. This is a big savings and cost and speeds up the process. Another aspect is the software tools are now bridging the gap between design and manufacturing. The design software is increasingly setting up the manufacturing process. Finally, the trend towards special collaborative communications software for engineers allows the team to efficiently collaborate at a distance.