Imagine slashing product development time—that's from initial sketch to mass production— by a third. At Guide Corp. (Anderson, IN), the largest automotive lighting company in North America, they've accomplished that. In just 12 months they've developed the ways and means to go from an 18-month timeframe to just 12. As they're involved in over 100 active development programs, this time savings is significant.
But the question, of course, is how?
There are two answers to that. One is to deploy cross-functional teams that focus all necessary resources on a problem until it is solved. The other is to employ a proprietary suite of predictive software tools that they've developed dubbed, not surprisingly, "Fast Forward."
Emphasis on Software
Several years ago Guide identified custom software as a key competitive tool. Rather than farming out simulation work or making do with existing packages, it put together a team of software engineers to make tools tailored to the needs of lighting development. This team has developed what the people at Guide—based on benchmarking and customer feedback—think is the most sophisticated software for their specialty in the world. According to Jeffrey Mickel, executive vice president for Engineering and Development at Guide, Fast Forward can use simulations to validate the optical performance of designs with over 98% accuracy–eliminating the time and expense associated with early prototypes. "The whole point of having these computer tools is to have confidence in what we simulate on the computer and not have to make a mock-up," says Mickel.
The first tool Guide uses is a packaging evaluation program called MaXpac. (Someone at Guide has a penchant for the term "Max.") The software can be loaded onto a laptop and taken directly to an OEM's design studio. It provides instant feedback to stylists on whether the shape and size of a lighting design they are proposing are feasible when it comes to optical and thermal performance. Once surface data from a clay model or a sketch is input, stylists can select the lens and reflector materials, paint finish, bulb type, and optics pattern necessary to get the look they are trying to achieve. The program immediately calculates the initial feasibility of the design, so the design team does not head down a dead end. If the design exceeds the parameters set by the program it is given a red light. After adjusting the criteria, say changing a lens material that can't take the heat of the bulb to one with a higher melting point, a new iteration can be run immediately to judge the effect of the change.
With this system, OEM stylists can make some initial choices as to which features are most important to them and which they are willing to compromise on. Mickel points out that the purpose of the program is not to tell designers they can't do something, since this hardly engenders positive feelings, but to provide a range of alternatives that work and are cost-effective. He uses the development of the taillights for the General Motors 360 program (Trailblazer, Envoy, Bravada) as a case in point. Guide's software revealed that the GM design did not meet thermal performance standards, but the GM stylist was dead set on achieving a specific look. So, by using the program to modify some aspects of the lamp that didn't affect that look, Guide was able to give him what he wanted while improving thermal performance.
With a workable concept in hand, the engineers at Guide begin the detailed optical design and analysis using a tool called "OptiMax." OptiMax simulates the light from the bulb to be used and the 200 or so lenses, called "optics," that focus the light into a beam pattern. (Currently each of these optics must be modeled separately to create the requisite beam pattern, which takes a couple of days. But Guide is working on a way to input a beam pattern first and have each optic modeled automatically.) While other lighting makers use software with similar capabilities, Mickel says that the sophistication of Guide's simulations allow it to catch and remedy problems mathematically that others might miss before making a physical prototype. Thus, speeding the overall process. For example, OptiMax simulates not only the primary light source but reflections from the bulb's glass globe that can cause glare to reach levels prohibited by government regulations. Without this capability, prototypes might exhibit glare that simulations did not reveal, leading to a time-consuming process of trial and error masking to isolate the offending component.
Taking its simulations a step further, Guide has developed a computer-generated virtual driving scene that re-creates a night time drive down a half-mile of real road. This virtual test drive imports data from OptiMax or from actual lamp assemblies that have been examined in one of Guide's darkrooms. Side-by-side comparisons can be displayed on a split screen. Since most programs adopt an existing vehicle's performance as a benchmark, OEM engineers can quickly see if their performance parameters are being met. "This capability has been a big benefit in showing customers that our simulations are accurate and that they can make decisions based on it," says Richard Meyer, Guide's director of Optics & CAE. Given the immense amount of data necessary to create a virtual drive, it used to take about two weeks to put one together. Now, one can be assembled in a morning, burned onto a CD and sent directly to the customer.
Guide has organized its engineers into cross-functional teams that are dedicated to a specific project and remain with it from initial concept to product launch. This approach shuns both the idea of the virtual team whose members can change as a project evolves, and the traditional "silo" organization that places key responsibility with departments not programs. Mickel jokes that the most successful project managers under the silo system were those who sprung for the most coffee, since they had no direct managerial authority over their team members. But under Guide's current approach, the project team and its leader are given both authority and ownership. Mickel says this leads to a number of advantages for the customer. Responsibilities and contacts are clear from the outset. Analysis is cross-functional from the beginning of the project. So a mold filling engineer who might otherwise not be involved until the end of development can look for mold-related problems early on when they are cheaper and easier to fix. And continued engineering support throughout the life of the product is faster and more effective because the institutional knowledge is concentrated and easily accessible.
The cross-functional team approach also facilitates a rapid response process Guide calls "MaXAttack." This is an intensive design and visualization session that often occurs as a component of the early stages of product development, but can also come into play when a design must be changed quickly. Guide likens it to an emergency room's trauma team going to work on a patient, which in this case is an ailing OEM lighting design.
MaXAttack can shorten the initial visualization process from weeks of sending designs back and forth between OEM and supplier to a couple of days in which multiple iterations are created, reviewed and modified with everyone in the same room.
This capability has never been more necessary than now. Since GM is Guide's biggest customer by far and since Bob Lutz is making wholesale design changes in programs already in progress, executing quick design changes may become the order of the day.