It is time to include risk in your Supply Chain Optimization Model. Optimization software has yet to address the inclusion of risk issues within the solution. Optimization exercises are often theoretically based. The departing port and the arriving port nodes are probably in the model. Sometimes the shipper's factory to the receiving customer warehouse nodes are assigned. They might also include a designated 3rd party warehouse (often with only an assumption of where that warehouse is). Node assignment is a manual process today. Users tell the system the locations of these sites and the system models a solution for time and cost. A tolerance for time (say 30 days) is determined by finding the lowest cost that meets the required service level. The user then seeks a carrier who can meet that time/cost goal. Often, no analysis is done of the actual physical path the carrier might take with your shipping container - how it is actually routed.
In today's high-risk world, with diversion, piracy, damaged goods, and potentially unstable weather and geologic events, this lack of visibility could be putting you at higher levels of risk. Do your electronic parts ship from Singapore directly to Oakland? Or does the ship stop in Indonesia and then transfer your container onto a ship bound for the US? Most shippers don't know. Recent earthquakes and tsunamis delayed or endangered some shipments. And some shippers were surprised to discover their shipments were even in some risky ports.
Purchasing assignments not integrated with Logistics Network designs.
Long-term planning can often model a new supply chain design, and determine the best location for a new facility. But most often, we are working with models based on preexisting partners and facilities. Node assignment is also based on preset suppliers. In today's world we should reverse this approach. For example, users should define their needs (a part is needed), and then determine who is the optimal supplier of that part based on all the issues, including the logistics network. Today there are two different organizations that attempt to solve this problem. The eProcurement product directory searches the marketplace to find you that part. Later the logistics or planning department needs to determine an inventory strategy as well as a transportation strategy for that part. A saving on one side might be offset by more cost on the other, but we don't really know. The cost of the supply chain is often not included in the purchasing system, and when it is, it is often incorrect, since it is based on some methods that estimate. Transport costs are like the snow ball rolling down the hill, picking up more along the way.
The recommendation recently written by Bill McBeath on Total Sourcing cost methods also talks about assessing the real total cost of sourcing with complete transportation costs. That adds an even fuller dimension to assessing the total sourcing challenge.
Changing Compliance and Regulations
Keeping track of global compliance issues is a huge task. The complexities of these elements are also not included in optimization exercises. Right now, there is an extremely large docket of compliance and trade regulations in the US and China. If readers remember the turnover to the EU and the establishment of that trading block, there was a plethora of environmental, customs and trade regulations put in place. Our current Pacific trade routes could be similar.
Make it part of the normal business process to gain effectiveness and relevancy
Other more traditional risk planning issues include purchasing risks. The positioning of raw material at the early stages of planning is based on assumptions for demand. This is a problem that is well understood. But interestingly, modelers don't go back to make changes in the optimization model, when expected demand does not materialize. The change most often is reflected in forecasting and later purchasing systems. Most often, that assumption is not re-addressed in the supply chain optimization model. This makes usage of the model static. Static and disjointed planning systems often mean we are using mathematical approaches that don't align well with actual conditions, and therefore, will cause inaccurate planning. In addition, it will be hard to learn from the history of assumptions vs. actual when using disparate approaches.
One of the business practices that must be addressed, then, is the static nature of Network planning. This should not be an annual exercise. As more is learned and business conditions change, the elements which are dynamically changing such as real orders, supplier status, and energy costs, etc. change, the analysis engine should kick in with recommendations that prevent increased cost or losses in customer service. These often can be major changes. During the last energy crisis, when oil reached $140 a barrel, we discovered many of our clients had not thought to re-evaluate their networks. So we worked with them on risk-adjusted planning exercises.
Recommendations and Conclusions:
1. Take into account your theoretical as well as your actual transportation network - you could be putting a high-value process or strategic product at risk.
2. Optimization needs to include the risks associated with each supply chain.
3. Get advice or ensure your compliance departments are keeping up with trade regulations.
4. Build in assumptions, and have a regular and automated method to readdress those assumptions as market conditions change.
Solution providers to look at:
Savi Technology, Inc.
To view other articles from this issue of the brief, click here.