CAM and CAD?

We’ve always thought that gains in productivity come from changing work patterns, supported by investment, rather than as a direct result of introducing technology. But mechanical engineering design flow has been very slow to change.

Recently however, a significant organizational change has started, gained momentum, and is now routine. This is the outsourcing of manufacturing from large companies to small companies, and often from developed to emerging economies.

Observers usually describe this in negative terms, often saying that manufacturing jobs are exported from their country. It can certainly be a disaster for individuals stuck in industrial towns in West Virginia, Flanders, the Ruhr, or the West Riding of Yorkshire, partly because they are not getting enough help to cope with change. For many citizens of developed countries, however, the positive effect of change outweighs the negative issues and cost of help.

We all benefit from growth in the percentage of the worldwide population with disposable income. They are building a stake in peace and prosperity. Those with nothing have no stake. The vast disparity in wealth that exists today is unsustainable.

The developed economies benefit from increasing numbers of potential customers in emerging economies wanting their sophisticated goods. Enterprises and individuals who can adapt, re-skill, and reorganize will get the benefit of outsourcing, even if they are located in the outsourcing (and more developed) economy.

The effect of outsourcing is that the disciplines of design engineering, supported by CAD, and production engineering, supported by CAM, are increasingly taking place in different companies in separate locations. We see declining machine tool sales in the developed economies and declining Computer Aided Manufacturing (CAM) revenues in North America and Western Europe, while sales are growing in the rest of the world.

Up to 25 years ago, design engineering and production engineering were separate. Even when the engineers were in the same company on the same site, they often had minimal communication with each other. With the introduction of multi-disciplinary teams and concurrent engineering, this changed. These teams took responsibility for all aspects of a product. Design for manufacturing and design for assembly were taken seriously.

While this coincided with the introduction of CAD, it did not in any way depend on CAD. Gradually, data transfer from CAD to CAM was established and improved. Mainstream CAD systems acquired CAM modules. More recently, users, especially those in emerging economies, have tended to purchase mid-range CAD systems, which do not have CAM modules, through resellers. The proportion of CAM revenues going to integrated CADCAM companies is going down, and the proportion going to CAM specialists is going up.

The main reason for this is that deployment of CAM systems is much more complex than the deployment of CAD systems. As a result, resellers of CAD systems have found it difficult to make money from CAM. There are several factors involved. First, greater know-how of the manufacturing process is needed to support a part programmer using a CAM product. Second, tool paths have to be converted into instructions for a specific machine tool via some kind of post processor. And third, the CAM system has to be physically linked to the machine tool. Few CAD resellers have a song, a smile, and a soldering iron in their briefcase.

More of the added value of a CAM system is in the deployment of the system rather than in the product itself, which is quite different than CAD systems. As a result, the CAM market remains highly fragmented, with many software developers and many specialist resellers.

Modern machine tools provide lower cost parts production. There are more functions to avoid wasted time in set ups, to handle material automatically, to monitor tool wear and replacement, and to reduce environmental impact. Machine tool controllers and communications have improved. They are not yet ‘plug and play’ Windows certified, but do contain standard computing components, whereas ten years ago they were custom designed electronics. They can be relatively easily integrated with factory networks, Manufacturing Execution Systems, and ERP systems. But CAM systems need capability to harness these functions and provide more throughput, albeit at a somewhat higher price.

Most CAD systems provide little or no support for a design for manufacture initiative. They persist in recording a single view of the geometric definition of a part. That is simply wrong when manufacturing processes and production-engineering constraints are taken into account. For example, a pocket will have a draft angle added by the tool cutting it. When we model the final geometry of the part, we are not modeling the geometry we want to send from CAD to CAM. There are still problems with surfaces. Elements with small slivers or sharp points have ill defined offsets. The CAM system user may have to remodel or make assumptions to calculate the offset geometry for machine tool paths.

When the production engineer prepares a process plan, machining is only one part of the process. In addition, there are material handling, set up, finishing, and assembly. Computer Aided Process Planning (CAPP) software helps to automate and improve plans by retrieving past plans. However, their use is not widespread. Historically, the main barrier to use was the additional effort required to code and file each plan so it could be recognized and retrieved when a similar part needed a process plan. It was often quicker to start from scratch. However, recent developments in unstructured searching, spurred by the Internet, look promising in helping to retrieve plans. A new generation of CAPP solutions is emerging from firms like Montreal’s Polyplan.

Outsourcing as a business initiative, further separating production engineering from design engineering, is unstoppable. The conclusion is that we have to examine the resulting design flow, identify the issues and opportunities, and then improve and adapt to the new circumstances.

The danger is that as production engineering moves to other countries, manufacturing know-how will be lost and mechanical engineers will create designs that are more difficult and expensive to manufacture. In the long term, if manufacturing know-how is not retained, design will relocate to where manufacturing is sited.

However, this is not necessarily going to happen. A design flow running across geographically distant locations could be tolerable. Engineers’ education would need to include design flow, production planning, and manufacturing, as well as design. Prototypes created during design, even if they are virtual, can incorporate manufacturing knowledge.

There will always be a need for more technology, but it won’t be in the form of improved CAD tools. In addition, there will be a need for more advisors—software tools similar to those that verify model integrity. Software could be developed to examine part designs for manufacturability, suggest easier tooling, and perhaps materials. But the major requirement is a tool set to improve communications between remote sites.