Pyramid Operating Systems
Pyramid's current CEO and co-founder, Gene Kirila, started Pyramid in 1984 as a manufacturer of fitness and rehabilitation equipment. It soon became known worldwide for making the "Mercedes-Benz" of that industry. The manufacturing process was world-class, too. In 1993, Pyramid applied for the Malcolm Baldrige National Quality Award, and the application merited a site visit.
Exposure to high-performance methods led Pyramid to question the unpredictable variance experienced in composite materials and components. In 1989, Kirila and his partner, Jim Lapikas, established a molding team to look into design process standards for a new composite operating system.
The team soon found that established methods for manufacturing thermoset composites everywhere in the world suffered from an ugly combination of faults. They were long leadtime, labor-intensive, capital-intensive, difficult to control, inflexible, and environmentally hazardous. If that weren't enough, the margins in the industry were also low. Seeing a great opportunity, the team took a clean sheet, science-based approach to narrow the variance of thermoset process, cut the capital and time required, and cleaned up the environmental problems, too.
After eight years and $20 million investment, the project finally culminated in a commercial process in 1996. That year, Pyramid opened the first VEC production facility on their 15-acre campus in Greenville, PA, northwest of Pittsburgh, where 65,000 ft. are under roof. Today the campus houses five units: corporate offices, mold shop, and open mold shop for pattern making, production plant, and chemical mix plant. You can check their site www.pyramidos.com
Pyramid has compounded two breakthroughs into a winning formula that they call Virtual Engineered Composites or VEC. The first part of this double whammy is a major advance in molding thermoset plastic composites-the core VEC process featuring a floating mold. The second part is the rapid deployment of the VEC operating system, called VEC 5.5, to industrial customers. In effect, Pyramid is a "process franchiser" to customers who are tied to Pyramid's center of process excellence by a telecommunications umbilical cord. Pyramid personnel don't have to be on-site to collaborate with the customer.
Pyramid's operating system differs from standard turnkey offerings because Pyramid monitors real-time operating data from remote VEC cells and maintains a round-the-clock, seven-day-a-week Solutions Center. Customers using a VEC cell can ask for an advanced data analysis or for consultation on a problem at any time. Service reps in the solutions center can see (and analyze and record) the same data as the customer. They can multiply the expertise by splicing anyone at Pyramid into a conference with a customer. For problems at the fringe of existing technology, Pyramid can tap into a network of outside specialists like polymer chemists at universities. The system deploys a best-in-class process to customers with a minimum of personnel and process develpment on their part.
The type of part commonly used to describe the VEC system is a "fiberglass" boat hull. The oldest method for making such parts is "spray up" on an open mold. The main advantage of this method is its flexibility. Pull out any mold, lay up the reinforcement, mix the resin, and spray.
But open spray molding has many disadvantages. It is labor intensive. The resin formulation has to be adjusted to the current ambient conditions, which vary because of ventilation for the fumes to escape. Consequently, the process is subjectively controlled and the work is not highly desirable.
To escape these problems, closed molding systems have become more common today. Molding compounds, either in sheet form or bulk, are typically made several days before use. However, the chemistry of uncured compounds is always active, so blending must allow for changes in the gel, cure, and rheological properties of the compound during the time before use. The molds themselves are generally major capital items made from expensive alloys.
The objective of VEC was to overcome the worst features of the established systems. To do this, the heart of VEC is a floating mold, plus instrumentation.
A master model is built to create the female and male mold surfaces, or skins, side a and side b. Necessary sprues and ejectors are designed into the mold skins. Each of the two skins attaches to its own mold cavity using a custom, fluid-tight flange. Any part feasible to mold that can be fitted inside a mold cavity can be molded. Once the enclosed cavities behind the skins are filled with a non-compressible fluid, like water, the mold is ready for use. Starting from artwork, a mold can usually be ready for use in four weeks-half the time of open face, and a tenth the time of a traditional hard mold. In use, molds can be exchanged in less than ten minutes-SMED setups. Floating molds greatly improve quality.
Process instrumentation is coupled with the ability to control the temperature of the fluid, so the time-temperature profiles of all molding cycles are better predicted and controlled. Consistency improves the outcome of each part, and reduces scrap rates and rework. Some parts can also be designed with fewer allowances for variance-thinner, lighter, and stronger.
The only performance measure on which traditional closed molding beats VEC is production rate. However, Kirila figures that traditional closed molding's disadvantages in cost, quality, and time-to-market make it tough to justify over VEC at any volume less than 10,000 parts per year.
Resin mixing is another major factor in part quality. The VEC system doesn't start mixing resins until just before the shot. Once a live brew of catalysts and additives starts to work, the molecules start a natural process of combination until the reaction is complete. Controlling this process is analogous to controlling the drying of paint.
Precision mixing just prior to the shot is necessary. In some cases, only a few drops of catalyst in several cubic feet of resin start the molecules squirming. To assure quality, the VEC system must monitor and control each shot, from mixing to removal from mold. The mixing ratios are adjusted according to the existent process conditions. The programmed sequence of steps is almost totally automated and fail-safe; the next step in the process cannot start until conditions show that the previous step has progressed to the proper point.
"Shots" are carefully controlled injections. A simple, forceful squirt into the mold won't do. Most parts contain a matrix of reinforcing material: Fiberglass, carbon fiber, or polymer fiber in a broad variety of densities and directional arrangements. The VEC system is capable of controlling a shot of resin so that the lay-up of any kind of reinforcement in the mold is not forced out of position. Injection pressure must not raise so high that mold skins deflect. For uniformity of cure, the chemistry of material entering the end of the shot differs from that which entered first. All material is injected in a single shot; multiple shots introduce unwanted variance into the process.
Not least, a patented air bleed sytem keeps bubbles of trapped air from forming in the piece. It also vents gases from the polymerization process itself. Control of the gaseous leakage allows VEC to easily pass environmental emissions standards.
To accomplish this, a huge database of process variables had to be assembled and validated. From it, simulations were developed. From the simulations, process control programs can be created for resin mixes and part shapes not processed before. Material and process characterization is advancing from sense-and-react to the parametric engineering level, which is a big step toward molecular-level engineering.
Using this system, each VEC cell at a customer site can be managed by one operator per shift. The on-site process control system links to the Pyramid Solutions Center. Software is written on a Microsoft NT platform so that data are also easily transportable into other software systems of the customer company. Data from each shot are stored, as on a flight recorder, and can be retrospectively analyzed.
The system walks the operator through the process using screens, check menus, voice reminders, and reminder lights. The VEC software-driven system replaces the paper operator instructions that were central to standard work. Operators (and others) can contact the Pyramid Solutions Center by phone, e-mail, and video conferencing.
Pyramid has five VEC cells in different versions operating at Greenville, making parts as a standard supplier of customers. Two more cells are placed with a customer, and five customer placements are in start up. The plant is a showcase and a training center for customers. Customer personnel typically come to Greenville for start up of a VEC cell, and when they are ready, it is transferred to the customer site.
Since Greenville is where VEC was developed, the workforce is multi-skilled. Operators who helped develop the process are not relegated to running production, but are also aides, developers, instructors, and troubleshooters-experts.
The Solutions Center is constantly collecting data on all VEC cells in actual use. Pyramid uses data to add to their knowledge of thermoset molding processes, so that Pyramid stays at the leading edge of actual practice.
Customers need not rely exclusively on their own expertise. They can draw on the full technical knowledge and history of the process collected to date. On the other hand, customers also limit the range of exclusive knowledge that they can acquire for their own competitive advantage, and some customers may be competitors of each other.
Pyramid has had no problem with "fire walls" between different customers' data. Customers sign on with the Pyramid VEC process once they conclude that thermoset molding processes are not a core competency essential to their business. Pyramid's primary qualification in selecting customers is to find those that have reached the conclusion that they can't keep up with thermoset technology, and so are willing to collaborate. In the case of a boat builder, for example, value is added by designing boats, not by compiling thermoset processing data, and a builder tied to VEC may be able to use thermoset engineering concepts that it would not have access to otherwise.
Distributed excellence is a major factor in the future of manufacturing. The cases of R-Theta and Pyramid integrate many current buzzwords: Internet strategy, customer success, outsourcing, manufacturing excellence, knowledge organization.
As the customer sees it, R-Theta is becoming an online design service whose software has credibility because it is directly tied to product realization and replication-a center of heat sink excellence. No customer has enough volume to afford to maintain such a center themselves. From R-Theta's viewpoint, excellent manufacturing and delivery have become a must-do- foundation of such a center of excellence.
On the other hand, Pyramid is becoming a center of process excellence for customers who want to either design using thermoset plastics or to replicate thermoset parts. In some cases the customer doing the design may not be the same as the one prelicating the parts. Few potential customers can justify trying to build their own thermoset center of excellence. Most, if they think about it, are not really in the thermoset business. Becoming deeply immersed in thermoset technology distracts them from adding their own intellectual value on behalf of customers. If you cannot afford the research to be at the leading edge of a technoogy, why not connect directly to a leading center for it?
Neither R-Theta nor Pyramid is self-sufficient in knowledge. R-Theta is forming ties with centers of advanced machining expertise. Pyramid has ties to polymer chemistry centers. They are becoming hubs that draw outside expertise into the subjects of their core competencies. In turn, they integrate if and package it for use by customers. In both cases, the expertise is embodied in artifacts shipped, but it is solving the customers' problems that adds value.
R-Theta and Pyramid remain very much in touch with reality. Their experts see, touch, and smell the materials and products daily. Therfore the design and process control software is unlikely to become "castles in the air"-guesswork models that poorly represent what really happens in a process. Reality gives the models value.
When Pyramid sets up a VEC cell on a customer's site, the hardware is only an enabler. The real value is the connectivity with Pyramid. Customers usually pay R-Theta for each sink shipped. However, R-Theta's value added to customers is only partly embodied in artifacts. The value growing at R-Theta is customer connectivity through R-Tools. Incidentally, both Pyramid and R-Theta have considerably improved the quality and efficiency of product replication, but that's a necessary part of offering the service-solving customers' problems.
Much outsourcing today is financially driven. Targeting activities to outsource may be as simplistic as shopping out the low margins and hogging the fat ones. Since margins from mass replication of artifacts usually stink, production is a favorite activity to outsource. However, the hope is that if a company offers a service that customers cannot afford to perform for themselves, they'll pay.
Piecing together a "service" based on margins from accounting reports is neither imaginative nor strategic. Rather, the key is visioning activities that from a customer's viewpoint the company does well or can learn to do well, and that customers, after their own analysis, would prefer to obtain from an expert. Then you can focus on your technology and solve you customers' problems better and faster-anywhere, anytime.
Concentrating on core expertise is the basis of creating a center of excellence, a new attractor for doing collaborative business through distributed excellence. Since none of us has perfect foresight, start with a promising combination. A center of excellence will improve through the learning self-generated through its own collaborative processes.
This concept of business is very different; success won't be instantaneous. Operating excellence is not achieved overnight. A world-class base of technology is not quickly developed. Communication standards and protocols can be a problem even with the Internet. Firewalls between competitive customers' projects may be an issue.
The basic idea is not new. Libraries, for instance, are often "centers of excellence." Patrons help themselves, but if they ask advice, they receive personalized service. Loose versions of this service concept can be seen in industries from Ready-Mix concrete to Venetian blinds. Most manufacturers, however, continue to pre-package their hard-gained expertise in artifacts and push them to market. Industrial customer-supplier relations today cannot be characterized as based on distributed excellence.
For manufacturing companies, one of the big shifts is redefining what has value and how value is exchanged. Most companies build the cost of service into prices charged for artifacts. Full-blown distributed excellence would more likely rely on service fees topped off by the cost of artifacts, much as shipping and handling are add-on charges today. Others might build the service fees into a leasing cost. In any case, the collaborative pioneers will have to draw many more partners into distributed excellence before one can discern a system unmixed with 20th century business customs.
Pyramid Operating Systems