Operations Network Design: A New Solution for an Old Problem
Tools for optimizing distribution networks have long been available, yet most of these applications still have an important limitation: they can’t fully model network requirements at the operations level. This limitation often leads to recommendations that cannot be implemented. What is needed is a new approach that enhances the current process by incorporating site-level detail and produces design recommendations that include real-world implementation plans.
The Problem With Traditional Approaches
Distribution Network Design is a strategic initiative impacting many important decisions, such as whether to open or close facilities, where to locate distribution centers, and which transportation modes to use. These projects generally follow a structured process consisting of three phases: data gathering, modeling & simulation, and implementation.
While this approach can help companies improve efficiency, it can also lead to recommendations that are costly or impossible to implement at an operations level. This happens because each phase of the process has inherent limitations.
During phase one, the design team gathers as much information as possible about the current network in order to understand how is operates. Using internal systems, interviews with key personnel, and other resources, they compile data on anything that has to do with distribution: costs, facilities, assets, customers, products, throughput, processes, etc. But despite their efforts, some information will always be missing, either because it was overlooked, was too hard to get, or simply wasn’t available.
The modeling & simulation step attempts to translate the data into a functioning model of the network. Despite increased processing power, however, it still isn’t possible to develop network models that include every level of detail. Typically, physical resources like drivers, tractors, trailers, dock doors, and warehouse staging area are not modeled explicitly; neither are customer specific service requirements and scheduling constraints. This is an issue because these factors can vary from site to site.
To cope with the data processing challenge, the design team aggregates the information they have. Values used to model a wide range of categories (shipment volume, customer demand, transportation costs, etc.) represent generalized totals that are not accurate at the site level. So, while it may be possible to develop a network model that is accurate at the macro level, the absence of site level details may lead to recommendations that are impractical to implement.
Fixing Network Design
Ultimately, the implementation phase is where the above limitations show up, as operations personnel struggle to act on design recommendations, or recognizing that some changes are impractical, fail to buy in to those that are valid. The net result is that most network design initiatives, while they may achieve positive results, generally fall short of accomplishing projected savings.
Operations Network Design (OND) is a new approach that provides a mechanism for evaluating network alternatives and testing the impact changes will have on individual customers. OND complements the standard approach, but is focused on a more limited set of variables, only considering the relationship between local network structures, specific customer service commitments, and available transportation resources.
By modeling distribution at a finer level of detail, OND makes it possible to replicate real world operating conditions more accurately and simulate network changes at the micro level. These changes can then be analyzed for feasibility, as well as for efficiency. This process allows operations personnel to fully vet strategic recommendations and gives them the ability to voice concerns during the design process.
OND is a plan-focused activity. As network changes are being simulated, a process for making those changes is also developed. As soon as the strategic planning and operations teams have reached agreement on proposed changes, an implementation plan is ready. This means action can be taken immediately and efficiencies can be gained more quickly. Examples for how OND can be applied include analyzing the impact of DC consolidation on specific customers or adjusting resource allocation during periods of increased demand.
Operations Network Design enhances design initiatives by introducing more specific details about individual locations. By including real world data, these initiatives ensure that strategic recommendations are feasible at the local level. The result are actionable plans that help companies achieve more of the promised benefits of distribution network design.