Feature-Based CAM Helps Produce Self-Guided Vehicle

In constructing an entry for a competition of self-guided vehicles, electronic systems developer DesignJug LLC’s (Atlanta, Georgia) Team Juggernaut had to machine complex parts in aluminum, steel and even wood. The challenge was that the team, which consisted of a diverse group of inventors, researchers, design engineers and students, had minimal experience as machinists. The parts the team needed to make, however, required professional-level programming and machining skills.

A feature-based CAM software package allowed the group to successfully produce the parts for its vehicle. The software, FeatureCam from Delcam USA (Salt Lake City, Utah), identifies individual part features and provides necessary machining parameters. These capabilities enabled the team to achieve accurate machining results. As virtual manufacturing enterprises and other cross-functional collaborations become more common on the U.S. manufacturing scene, the team’s experience may provide a glimpse of the future.

DesignJug LLC develops embedded, reproducible computing systems that can operate reliably in harsh and space-constrained environments while employing commercial off-the-shelf (COTS) components. Troy and Jean Takach founded the company in 2003, drawing upon two decades of experience in computing systems integration and electrical/mechanical design, creating monitoring, control and data-acquisition systems for defense, transportation and industrial applications.

The company specializes in “high-volume disruptive thought,” explains Mr. Takach. In other words, to attain the speed and quality of development its customers require, the company examines alternate ways to achieve effective results.

“The way that something was done 20 years ago shouldn’t necessarily be the way it is done today,” notes Mr. Takach. “Sometimes, the way you did it last year shouldn’t be the way you do it today.”

In mid-2004, DesignJug began to gather components and assemble a team to compete in The Defense Advanced Research Projects Agency (DARPA) Grand Challenge 2005. DARPA organized the contest following a Congressional mandate stating that, by 2015, one-third of all military vehicles must be able to operate autonomously. The agency created the competition in 2004 to accelerate the development of such vehicles. Twenty-five teams competed in the 2004 Grand Challenge, but none completed the DARPA-designed desert course.

“Such a nationwide competition has precedents and attracts a mixture of participants,” says Mr. Takach. “A contest is the reason Charles Lindberg crossed the ocean. An event such as this takes advantage of the competitive nature of Americans.”

Team Juggernaut consisted of students from the three Utah research universities, as well as business and technology professionals from across the United States. Included were software specialists, electrical and mechanical engineers, and project management and logistics experts. The team’s professional diversity reflected the company’s focus on doing things differently. The key to its success was translating creative thinking into mechanical reality. To put ideas into action, DesignJug employs a variety of tools and capabilities, including four-axis CNC machining, rapid prototyping and manufacturing systems, assembly, testing, welding, moldmaking, 3D modeling, system troubleshooting and advanced software development.

Intuitive but powerful manufacturing software was one of the tools that enabled the team to realize its ideas, says Mr. Takach. Companies want the ability to conceive a part and manufacture that part in their own facilities, he says. Using the software, DesignJug programmed the machining of structural components, enclosures and mounting devices.

As Mr. Takach articulates, the team members were not machinists, and therefore they did not know G codes. However, the software enabled the participants to concentrate on a part’s final application instead of on programming details.

For example, to program the part shown in the photo on page 138, the CAM software identified key features of a
mold for a plastic housing and then provided the tool paths and cutting parameters needed to machine those features. Team members subsequently machined the mold from wood using DesignJug’s vertical machining center from Haas Automation, Inc. (Oxnard, California).

“The CAM software is not a beginner’s version, but rather it is a full-featured package that ensures that those without extensive machining experience can be successful,” comments Mr. Takach. “Team members constructed parts with the software the same day that we introduced them to the package.”

Similarly, the company uses IronCAD (Atlanta, Georgia) solid modeling software because a user can learn the program in less than 1 day. Together with the shop’s fusion deposition modeling (3D printer) rapid prototyping machine from Stratasys Inc. (Eden Prairie, Minnesota), the software packages facilitate a fast-track production process. After a 3D model is created, the CAM software can export it in an STL format, which is then sent to the rapid prototyping machine. By using this machine in the middle of the process, the company says it can verify that the part is correct prior to machining.

“In addition to being easy to learn, the CAM software can quickly transfer from a CAD model into tool paths that we can place on our machine,” comments team member Neil Johnson, a mechanical engineering student at Brigham Young University. “We’ve used the software to mill parts of aluminum, steel and wood. Just the other day, we made a vacuum form mold for plastic out of plywood. We used the CAM software to export a model that we then carved out of the wood on the mill.”

The shop also makes molds for rubber gaskets and shock mounts using the rapid prototyping machine and STL files from FeatureCAM.

The company actually built three vehicles in preparation for the 2005 Grand Challenge—two ATVs and a four-wheel-drive truck. The ATV chosen for the competition progressed through the preliminary round and qualified for the final competition. However, an errant hay bale interfered with elements of the guidance system, and the vehicle didn’t complete the final course.

“The race proved our platform; we had no failures,” says Mr. Takach. “We just didn’t have enough time to implement all the software we needed.”

Because Team Juggernaut was a kind of virtual manufacturing enterprise, its experience with feature-based CAM software for machining may foretell the kinds of tools that will prevail in the future as U.S. manufacturing evolves. Much machining will be done by individuals who need quality machined parts as they collaborate with other partners in a virtual environment. Access to intuitive software and computer-automated machine tools will be essential in meeting this challenge, says Delcam USA.