Outsourced manufacturing is a key element in the playbook of just about every high-tech producer today. But it also presents some major obstacles to the forging of an efficient supply chain.
Radisys Corp. is a supplier of wireless infrastructure for the telecommunications, aerospace and defense sectors. Its supply chain touches the U.S., Canada, Asia, Europe and India.
Like many of its global counterparts, Oregon-based Radisys saw the benefits of outsourcing production. The savings in labor cost were evident, but they came at a price: an extended supply chain, with multiple outside partners, makes it difficult to plan for supply and demand.
Lisa Aleman, director of sales and operations planning at Radisys, joined the company in early 2010, just prior to the decision to take a captive manufacturing line and move it to Asia. But the new setup quickly ran into some complications.
“In the transition was the assumption that if you give a contract manufacturer your demand, that should be it,” says Aleman. “They’re just going to process it exactly as you would.”
Not exactly, as things turned out. Having shed responsibility for production, Radisys found itself lacking full visibility of supply and demand in the pipeline. As a result, safety stock residing within the system became static and unresponsive to the needs of the moment.
“Just because you give somebody a demand plan, it doesn’t mean they are going to produce and buy materials as you would,” says Aleman. Radisys didn’t know how the outside manufacturer actually acquired and dealt with those items. There was a substantial communications gap to be overcome. As a result, “we were not able to deliver to customers. And we were not able to give them information as to why.”
The need for automating supply and demand planning was obvious. Radisys had to get a handle on the tradeoff between fulfilling orders in a timely fashion and minimizing inventory cost. It also needed a forecasting tool that could support daily order commitments, in addition to the monthly cycle on which the old system ran. Certain aspects of the forecasting model had gone out of sync by as much as three months, leading to missed order commitments as well as excess inventory.
Search for a Vendor
The company looked at four or five software vendors, with an eye toward selecting one that could support a high-tech operation. It ended up settling on Icon-scm, which was subsequently acquired by E2open.
Simultaneously with the I.T. purchase, Radisys was adjusting business processes to accommodate it. The company was in the early stages of implementing sales and operations planning (S&OP), but had yet to realize the full benefits of that effort. Planning meetings were still being conducted by various departments in a siloed manner.
“None of that tied off in a holistic approach, with one plan, one owner and everything cascading off of that,” says Aleman. Eager to embrace a more integrated approach to supply and demand planning, the company “needed to move the business process ahead of the automation as much as the team could bear.”
“This was not a one-off software implementation,” recalls David Vallejo, a former employee of Icon-scm and now vice president of sales enablement with E2open. “It was really a transformation they had to go through, with many steps involved.”
Complicating matters was the fact that Radisys didn’t have a simple outsourcing setup. It was working with multiple contract manufacturers, and was even doing some final assembly in-house.
The E2open application gave Radisys visibility into the way manufacturers were handling orders, work in progress, and finished product ready to ship. It knows exactly which materials the partner purchased and when, as well as what’s available on the floor and in inventory. All of that data is now available on a daily basis.
For the first time, Radisys could adjust the flow of supply and inventory in line with present market conditions. If it accepted an order cancellation – a common occurrence, given that many of its customers condition their orders on winning business from their own accounts – Radisys could assess the immediate impact on materials. And if it landed a big deal, it could answer questions about product availability within minutes.
The system dovetails with Radisys’s S&OP process to allow for both long-range planning and immediate execution. It supports the monthly cadence that updates the company’s long-term planning, as well as the inevitable changes that occur out of cycle.
A Shift of Partners
Radisys ended up consolidating its manufacturing with one outsourcing provider. Aleman and her team worked closely with the entity to clean up the data, get it up and running on the application, and acquire the necessary visibility. Later, it shifted to another contract manufacturer because of a difference in the algorithms employed by the two original partners.
“We’re very low-volume, high-mix,” she says. “Their solution was high-volume, low-mix. The disconnect was understood, but they weren’t willing to modify their systems to what we needed. With the new partner, this whole package was negotiated in.” At last, the priorities between original equipment manufacturer and contracted provider were aligned.
In the process, Radisys has become better attuned to actual conditions in the marketplace. When Aleman joined the company, up to 90 percent of its product portfolio was being built to forecast. Since then, it has adopted a three-tiered model, in which about 10 percent is built to forecast, with the remainder divided up between vendor-managed inventory arrangement for materials that aren’t yet committed to finished goods, and a true build-to-order process.
As a result, Radisys no longer finds itself in the position of bringing in materials for orders that never drop. Even safety stocks are mostly in the form of semi-finished items, adaptable to the needs of multiple customers.
Radisys was able to automate the majority of its supply chain within about six months. Along the way, it has slashed the cost of excess materials through better procurement policies, while making improvements to forecasting and product engineering. Sales, operations and finance have finally been aligned, and customer lead times have been reduced.
Looking ahead, Aleman expects to bring more strategic focus to the S&OP process by addressing the demand side, a more recent initiative. It also wants to prioritize orders “so that we don’t have lower-tier customers stealing from higher-tier customers that are actually forecasting.”
Also on the agenda is the ability to collaborate with upstream suppliers that deliver to the contract manufacturer – the goal, in Vallejo’s words, of “multi-tier process orchestration.”