Inspiration for Hardware Design
Inspiration for Hardware Design
MTS Test Systems had a number of challenges, including decreasing profitability and increasing lead times; a proliferation of products and parts that was scaling with growth; and competitors that were offering “good enough” products with lower prices and shorter lead times.
MTS decided it needed a transformative change and launched a modularity program that delivered significant results, including:
Before modularity MTS used 11,000 parts to build 150 product variants. After modularity, the company used 800 parts for 100,000 product variants.
MTS Systems is a global supplier of test systems and industrial position sensors for a wide array of applications, including wind turbines.
MTS Systems had a number of challenges, including decreasing profitability and increasing lead times; a proliferation of products and parts that was scaling with growth; and competitors that were offering “good enough” products with lower prices and shorter lead times. MTS decided on a transformative change.
Beginning in about 2002, market growth for test equipment was fueled heavily by newly developing countries where local companies were starting to design and manufacture higher quality, higher complexity products. MTS, who had established themselves in these markets years prior, was well positioned to take advantage of these new opportunities.
As leader of this niche market and inventor of the core technologies, MTS had seen many years of strong profitability. Customers were willing to wait a long time and pay a lot of money to buy a system from MTS. As new competitors entered the market and existing competitors chose new ways to compete, a new level of products was introduced with “good-enough” performance, lower price and shorter lead times. Some customers could no longer justify the additional time and money required to purchase from MTS.
The company was organized into groups around small market niches that included automotive shock absorbers, aerospace aluminum, asphalt road materials, artificial knees and many others. Each group managed a separate profit-and-loss perspective and controlled profitability of every order using margin-based pricing. Over the years, this arrangement led to the branching of MTS’s core business approach toward the optimization of price performance for the narrow slices of the market. It also led to the proliferation of product variations within product areas that required an increasing number of parts to produce. Operations had a very difficult time gaining leverage, and it was hard to identify opportunities for improvement to the overall business.
Continued growth in the market was limited by MTS’s ability to address the demand for lower prices and shorter lead times. Their project-based operational model became ineffective resulting in increased cost, lead time and quality issues. Operations were pushed to capacity, and the scaling-up of the current system was not a viable business option. A new approach was required to meet the broad range of customer requirements while maintaining long-term profitability.
A small, disconnected marketing team at MTS attempted to manage the full breadth of markets, products and customer applications that formed the niches. Each area generated new requirements and potential changes to the product, but it could be five or more years before development resources were available to make changes. It was very difficult to keep products fresh once they were launched.
Sales teams were organized by niches, and they consisted of both sales and application engineers. The teams were capable of gaining a deep understanding of what each individual customer needed and would go out of their way to deliver exactly that. Their pursuits were sometimes detrimental to the company in how they consumed time and resources. The teams often knew more about an application than the customer and needed to coax decisions in order to configure a test system. The consequence of this structure was that it led to a broad range of different product configurations. Test systems were built for additional customers within niches, but there was almost no repetition across niches.
In order to improve the efficiency of selling, a product configurator was created for a few product areas that included a range of product performance, features and options. It greatly improved the accuracy of the communication between sales and engineering, but it made no improvement to operations or product management. The configurator included all reasonable product properties that customers may want, but there was no predetermination as how the product would be produced, if it had been built before, or if it was a profitable configuration. Since there was no active connection between the sales configurator and operations, sales people configured whatever the customer thought they needed, and engineering decided how it was designed and manufactured.
The product design function at MTS was closely coupled to project engineering. Each customer project had an engineer assigned to follow the order through the system and to troubleshoot when something went wrong. The planning of projects always included a contingency for schedule and cost both of which were burdened by the customer. The goal was to ensure project profitability and deliver it on a promised date even if the date was later than what the customer originally requested.
Each customer project was accompanied by a new set of paper documentation. Little efficiency was gained in engineering even when the sales teams were able to sell close copies. After customer projects were completed and quality problems were resolved, the design teams could then focus their efforts on new product development. Consequently, development activities took a back seat and they were risky activities for product teams to propose. Most of the innovation and new designs occurred on customer orders with the risk built into the price and schedule.
Efficiency gains in engineering were realized only with individual people. Based on their own experiences, each would develop processes to optimize their efforts. In many cases, it was easier for individuals to start from scratch on a project versus reusing someone else’s documentation. This mode of operation covered-up the system-wide challenge because people were doing their jobs well. Problems occurred only when the system of people was disrupted.
Since every customer test system was new to some extent, many issues during manufacturing became science projects. Each took a lot of time and brainpower to resolve, and involved highly skilled and experienced people. The learnings from issue resolution were stored in the heads of the people who worked on the problem because there was no process to share across the company. Project engineers would try to pick the same operational resources each time they needed to build a system in the same product area.
The wide range of components supported by the supply chain was managed without the support of forecasts or the ability to buy anything in volume. The lead time for projects was established by the component with the longest lead time, and the low volumes offered MTS little buying power and control over prices or designs. Components that were available one year may not be available the next year.
In 2002, MTS was focused on improving the efficiency of their operations by reducing the effort for individual projects. They needed to meet the lower cost levels required by the evolving market. The organization was over-burdened and stressed, but the business proposition to scale-up operations in its current state was not attractive. A new approach was needed to regain control of the system and meet the changing requirements within the market.
These operational issues along with a desire to control and improve overall profitability caused MTS to make sweeping changes to the organization. With a new organization aligned by business functions, they hoped to shift the focus to core profitability and to the identification of business-wide opportunities. As a result, the sub-optimization within market niches was gone, but the pendulum had swung too far in the other direction. Specialized knowledge was not being delivered where it was needed in a reasonable amount of time.
In 2004, MTS launched a series of initiatives to balance the organization and secure their ongoing leadership position in the market. Called the Market Leader Challenge, it sought to improve the competitiveness of the company’s products and processes. Using modular architecture to efficiently deliver products was supporting the Leverage Product Offerings initiative. It would create a new operating model for higher volume products to keep up with market growth in emerging regions, and it would expand the range of product performance and features to address new product entrants.
MTS chose a single product family to pilot modular architecture and intended to work on others in succession. A pilot approach was chosen because the business systems and process that were being created for the product family were not replacing existing ones. The new process could be developed and tested on a small scale with new people and then repeated within another product area. The management team picked product areas that reached across multiple niches and consolidate technical capabilities, and chose servohydraulic load frames as the first.
Servohydraulic load frames are used by MTS customers for complex fatigue life and fracture growth studies on metals and composites or for conducting simple tension and compression tests on consumer products and building materials. This is a market niche where competition had increased and MTS market growth potential was threatened. To maintain a competitive offering, they sought to reduce customer lead times significantly lower than their 48-day average.
Lead time reduction would be accomplished though the new modular architecture that would collect the full scope of this product area into a single product platform. A goal was set to achieve 80% of orders using product configurations of 100% established modules. These could be delivered in a fraction of the 48 days. The remaining 20% of orders would use 80% established modules and would reduce the time to both design and deliver.
The management team determined that the additional capacity required to meet the growing market demand would come from increased efficiency rather than the addition of resources and manufacturing space. A large chunk of the efficiency would come from having more orders configured instead of designed, and this would lead to a 30% reduction in job cost.
The operations would also become more efficient by reducing the number of part numbers that needed to be managed by the various teams. By creating a modular architecture, a 90% reduction in total part count was expected. MTS also established a goal to keep new part creation to less than ten per week.
With a new approach to delivering solutions to the market, MTS has added the efficiency and capacity needed to keep up with the market growth. By leveraging the similarities of the test systems across applications though a deliberate family of products, they were able to take advantage of some level of scale. In the words of CEO Laura Hamilton, “Modular platforms have helped us build a cost advantage. We’ve reduced our unique part numbers.”
Much of the success at MTS was due to the implementation of teams within the major business functions. These teams of managers and core implementers worked together to figure out how to implement the Modular Architecture in their function. The successful results encouraged the management team at MTS to continue to develop Modular Architectures for other products. This included the family of electronic control system and the family of software applications.
The initial launch of the servohydraulic load frame product family was a great success. It incorporated an extensive update of the products across the many application niches that would have individually taken ten years or more. The number of overall product variants was increased by more than 100%. Features and system performance for each of the individual applications was increased by sharing across the entire product family.
At the same time the new Modular Architecture was being developed, the sales organization was changing to support the Market Leader Challenge. Sales teams were being deployed regionally instead of by market or application. They were becoming generalists instead of specialists and needed to rely on others for specific application knowledge. The ability to configure products from a predetermined set of designed and managed variants was an important factor in their success with the change.
A new product configurator was implemented that was linked directly to the operational systems. It was based on the modules within the architecture, and whatever was sold by the sales team from the configurator was built by the manufacturing team. The efficiency of the product family was controlled with the predetermine variants offered in the configurator. The people-dependency of delivering product was removed from the process.
A more efficient marketing organization also evolved out of the development of the Modular Architecture. A single product manager was assigned to the management of the common platform components that made up the core of any system. This individual is focused primarily on managing the internal process and issues to improve the operational efficiency of the product. Other market and product managers are focused on the components like software and test fixtures that adapt the core of the system to specific applications. These folks can manage multiple applications and limit their involvement in the core product.
Engineering activities at MTS have increasingly shifted away from delivery of specific customer orders to the up-front design of a planned product platform. In fact, the development of the first Modular Architecture for servohydraulic load frames solidified the need for a research and development department. MTS found that it was necessary to separate the engineers working on new products away from the demands of everyday customer projects.
The platform design of the load frames is a marvel of design efficiency. The overall part numbers were reduced by 75%- 90%. Before the modular architecture there were 11,000 unique part numbers to build 150 unique product variants. With it, they only need 800 unique part numbers to build over 100,000 unique product variants. Many customized test systems can also use a portion of the platform integrating newly engineered content through standardized interfaces.
Documentation is one of the key engineering functions that was addressed in the load frame project. It was once a paper-based process where project-specific drawings were created, printed and shipped with each order. It is now a paper-less process that automatically builds CAD models and simplified eDrawings from product configuration data. It eliminates much of the non-value-added time engineers used to spend on each order, and individual project costs have been reduced by 18-35%.
The electronic documentation now also contains all of the manufacturing information. When a product is configured, the production operations are configured at the same time. The documentation has also become an important tool for governing the Product Architecture. Changes now needed to be planned and approved. Using paper would have allowed the organization to go back to its old ways.
The initial Modular Architecture project proved to the management team at MTS that they were on the right path to regain capacity and efficiency in their operations. For load frames, the average time in the manufacturing process was reduced from 48 days to 14 days, and the corresponding work-in-progress was also reduced by 75%. In the past, systems would travel upwards of 1.2 kilometers in the plant. Consolidation of operations into modules and the implementation of an assembly cell reduced travel to less than 100 meters. Modules are built and tested ahead of time to ensure they are working when assembled into the test system. The number of steps in the assembly process was reduced from 24 to 10, and there was much less rework.
The reuse of components across the broad product family allowed the supply chain to be managed with forecasts and higher volume purchasing. The team established tiers of components to go along with their planned product lead times. Tier I items are high demand and closely managed to be always available with short lead times. Tier II items have slightly longer lead times that expand with high demand at a single point in time. Tier III items have the longest lead-times and are not typically held in inventory. These lowest volume components are procured to support each order, but the lead time is consistent.
A significant set of modules that was developed for the modular architecture within the load frame product platform was the integrated actuator beam. Analysis of the functions of the components indicated that many of them could be combined into this super-sized module. This component is being built and fully tested before it is integrated into the rest of the load frame.
The standardized interfaces surrounding this set of modules allow it to be integrated in multiple configurations of the load frame. Much of the variation required in the system is accomplished with a separate, bolt-on hydraulic manifold module. Variants of the beam module are built using late differentiation. The long lead item of the casting is produced in a limited number of sizes and stored in inventory. The casting is machined per unique variant as it is demanded by customer orders.
This set of modules did not result from the work of a clever design team. It was a result of deploying the Modular Function Deployment® method to identify the location of module interfaces that would accomplish both functional and strategic objectives of the product. The significance of the design is measured by the impact to the customer and the business as a whole.