Where Is the Software Industry Taking Us?

In the constantly changing world of software, one thing is certain: very few things remain the same. The reason this particular world is in a constant state of flux is because the industry is a beast shaped by developments not easily visible to the rest of us. Not surprisingly, many of these developments have sneaked up on both customers and software companies alike. Having any change take place in such a stealth mode can be unnerving.

Most individuals and companies that were involved in the PC movement of the 1970s knew that computers would one day be popular, but not even those with the Midas touch could have predicted the PC golden age. Another example is the rise of e-business, which has forced software vendors to accelerate the evolution of many integration software products, customer-facing applications and various outsourcing schemes. New tools and services have emerged in the business-to-business and business-to-consumer areas: new portals, trading exchanges and enterprise application integration.

Application service providers are now popular with companies as time-to-market, complexity of technology and an information technology skill shortage puts a stranglehold on the market. Until now, makers of specialized industry-specific applications were limited by geography and peoplepower. But with the rise of the Web, companies can peddle their (soft) wares to anybody in the world because high-speed Internet connections make this a reality.

"It is a given that the Internet, be it wired or wireless, will continue to play a major role in the development of new technology," says Dan Marinac, vice president of marketing for Cimatron Technologies (Burlington, Ontario). "The internet is a fast growing technology. Many companies are looking to the internet/intranet to improve communication and, as a result, efficiency. We are finding that more companies would like to publish their process information to the intranet, through the use of graphical setup sheets that makes life easier for the operators and reduces the likelihood of errors."

But besides Web access, just where will software trends take design engineers and product developers in the future? How will jobs change, product development be influenced and time-to-market be affected? And of these trends, how many will have the greatest impact? Cell phones, netsourcing, laptop computers and hand-held computers are just some examples of how current software trends could take the industry in an entirely new direction. In 1997, handheld computers were considered expensive, high-tech toys, but as technology continued advancing, so did the uses for these devices. They now have become powerful tools that owe a great debt to the power of software.

Although the terms CAM and CAD occasionally appear side-by-side, transferring parts files between the machining stage and the design stage of the manufacturing process has never been as straightforward as it could have been. Engineers have the ability to model any part on a 3-D CAD system, but what they don't have is the ability to exchange their models with vendors, customers and colleagues, who use different CAD/CAM systems for editing, design-for-manufacturing and manufacturing. The use of pure geometry such as that supported by industry standard data translations such as IGES, STEP, SAT and parasolids does not allow users the ability to modify solid models that lack parameters or features, as well as history-based solid models. As a result, teams cannot work on design projects together no matter what underlying CAD software is used. However, some CAM software peddlers do see a light at the end of the tunnel - a single source for CAD/CAM - where applications share common attributes from the solid.

"For design engineers in the near future, there is going to be a tighter integration between the design model and the machine model," says Ben Mund, marketing manager at CNC Software, Inc., (Tolland, CT). "Most CAM systems today are able to accept a wide variety of part data, but many are still limited in their capacities. Soon, it will be very easy for design engineers to make changes and make sure that those changes are made correctly, so they don't lose work they did in the prototyping process. This will have the potential to speed up the process because a lot of the work that goes to programming will not have to be duplicated."

Mund says high-speed machining (HSM) will become a more prevalent trend in part because it's an efficient way of doing complex mold work. And since the HSM hardware is becoming more affordable, the industry will see reduced production times across the board. In addition, as designing and machining come closer together, compression of a parts' time-to-market will only increase.

"We are focusing on HSM pretty aggressively, recently creating a utility in one of our software packages that can vary the feedrates," says Mund. "We programmed in a function that looks at a model and, based on the volume being cut, can slow down or increase the feedrates. If it's a shallow stock the rates increase, if it's a deep stock the software will slow the feedrates down. This extends the life of the machine tool cutter and reduces cut time. It will be systems like these that you will see doing more complex work in shorter time spans."

Forecasting what developments will likely take off Mund says is hard because no one knows exactly what the competition is up to or coming up with. But CNC is at work developing broad capabilities that will allow any type of data to be "transparent" to the users so they can utilize it on anything CAM-related. The software will have better stock recognition - meaning that an engineer will be able to run a toolpath and have the software understand what is left over in order to generate a toolpath to those areas.

"From a CAM perspective, we'll be able to tie a number of tasks together in the next few years," says David Boucher, director of development at Pathtrace Engineering Systems, Inc. (Berkshire, England) "For example, five years ago everything was designed using wire frame, but now we are experiencing a fast move to models. Those on the shop floor will be able to design parts that are very similar by entering just one or two key dimensions to regenerate it - in other words, they'll be working with automatic feature recognition. That will tie the design and manufacturing process closer together so things will move faster."

Boucher believes that shortly there will be a time when manufacturing information will be an integral part of the design, with the CAM process automatically retrieving that information. He also envisions integration between different third-party applications, and the sharing of common information. For example, there may be information in the solid model that both the CAM and the FEA systems use - downstream applications sharing common attributes from the solid.

Boucher also believes that the Internet will have a huge impact on the shop floor in the coming years - bigger than the effect it's having now. Since the Web has proven itself such a good vehicle for sharing information between companies, he thinks it will have a strong impact on improving CAM tool machinery in particular because it will allow more mills and lathes to be used together, instead of having to shuttle back and forth between machines.

"CAM tends to be a conservative field, often following CAD in its new innovations. But we are seeing a strong drive between the solids and the Internet," Boucher says. "Large models had maybe 100 surfaces just a few years ago, but today models commonly have about 2,000 surfaces. This puts a huge strain on the computational ability of the CAM software. We are, however, seeing a trend toward increasing utilization of software for manufacturing processes. So, these issues will eventually be overcome."

Well, if pictures are worth a thousand words, then what are 3-D CAD model images worth to a company trying to communicate engineering designs to designers at another business developing their product?

That net worth, companies will tell you, can be incalculable because if the design data does not translate well or if updates don't make it into the final model you have big problems. This scenario has happened to more than one design firm throughout the years. What many in the CAD world have hoped for is a kind of "big bang" where someone comes up with the foolproof, streamline system that automatically renders changes and is able to translate itself universally without needing any specific functions. But the big bang is more of a wish than a reality. The CAD advances have come as a result of evolution through inventive incrimination, not a sudden flash of brilliance.

"I do see jobs changing rather dramatically with software's advancement," says Rishi Madabushi, software designer for IBM (New York, NY). "Engineers will be freed to be more creative and thereby create more innovative products and enjoy what they do a lot more. Database management systems will one day take charge so engineers won't have to worry about the clutter of what work belongs to whom. Knowledge management systems will capture things as they are created so engineers won't have to come back and redo things.

Generative systems will show the design team how long a part will take to create. So, what we are going to see is better specialization, creativity and innovation."

Madabushi says that knowledge management is the most critical thing the industry is chasing right now. Sending flat 2-D drawings or 3-D CAD models without product knowledge does not provide an engineer much value. But knowledge management can be crucial from a CAD/CAM and NC verification perspective in terms of driving the above information: Do they know how much it weighs; do they know what it is made of or who the potential suppliers are? Little things like tolerance factors can alter a company's entire supply chain philosophy.

Madabushi adds that there also will be software products set up in the near future to solve problems for specific industries such as automotive and aerospace: a standard set of generic CAD applications versus the present way of solving problems through bottoms-up modeling. In addition, there will be a lot more software adaptation to other systems where people can pick up and customize their application to suit their specific needs. Directly linked to all of this is the concept of the open marketplace and the sharing of product content between companies. Prior to this, companies had to go to whomever they could find locally. Now, companies are able to work with the best prices and materials possible. Some of this extended enterprise is already going on at a high level. Standards concerning company data exchange also are evolving at a rapid rate.

"I see a lot of potential, but I don't see where it's going to take us because the technology is not as mature as it could be," says Madabushi. "I don't know how far off this stuff is and I would not even hazard a guess, but there is a lot of potential out there. Twenty years ago if you talked to aircraft engineers, the objective was to build the most efficient airplane," says Madabushi. "Now their concern is 'are the reading lights bright enough and will the toilet seat stay up in turbulence.' The barriers have been taken away and flying is just a given. Pretty soon with software, this will all be a given as well."

Engineers, industrial designers and researchers are working with million-dollar virtual reality (VR) systems to create visualization systems that let collaborators see and manipulate 3-D models of their work projected in space. Visualization systems are typically used to view computer-generated models of automobiles and other product prototypes or 3-D representations of complex research data. It is no longer important just to visualize a single part or even look at it in 3-D, but to look at it in the concept of the whole part.

Engineers don't want to visualize but to touch, tap and see what sort of metal is being used. In other words, engineers are moving away from the capture of data to the use of data for decision processes and analysis.

"The major trend in VR is toward collaborative visualization of 1:1 scale products and projects both locally [disparate groups work together on product development in the same room simultaneously - a form of concurrent engineering] and remotely [product teams in different geographic locations interact simultaneously with the same virtual model]," says Linda Jacobson, virtual reality evangelist and senior manager with SGI (Mountain Lake, CA). "The result: better, faster decision-making and insights. This collaborative visualization capability will rapidly accelerate time-to-market, and reduce costs and enhance quality in product development. The use of collaborative visualization will require people who understand how to develop and/or drive real-time 3-D graphics applications, and who are familiar with advanced visualization interfaces - both software and hardware."

Jacobson says that the biggest software developments she sees coming down the pike will be improved support for multi-channel stereoscopic displays, and improved support for collaborative environments. SGI is presently developing lower-cost, easier-to-use collaborative visualization displays, software toolkits to facilitate creation of VR applications and R&D of collaborative visualization of software products. "Creating data from software is easy," says Jacobson. "making it useful is the difficult part."