Distribution ERP and supply chain management

Not long after MRP (Materials Requirements Planning) was commercialized in the 1980s, DRP (Distribution Requirements Planning) emerged. MRP was seen as 'the next big thing' and a panacea for the industry's problems so, when somebody realized that a bill of material turned on its side looked like a supply chain, using MRP's logic as a supply chain management tool took off because companies saw the value in taking cost out of the supply chain and now just passing it back. In more recent times, supply chain management (SCM) has emerged as a management technique in its own right so DRP has somewhat faded into the background: it is still there but it just doesn't get the same attention it once did.

It can be argued that this doesn't actually matter because dedicated supply chain management systems have matured and now satisfy this market, but that ignores two things. One is that the requirements of many companies may be satisfied by the functionality that mainstream ERP packages offer and these may be cheaper than some specialist SCM systems, and secondly that in the world of ERP things are frequently not black and white, so many manufacturing companies that need ERP also have their own trade or retail distribution chains in addition to their regular ex-factory sales and running two separate systems may, therefore, be neither advisable nor indeed necessary.

So, if a manufacturing company also has distribution outlets, can it use its regular ERP system to cover DRP also? The good news is, yes it can. Going back to the bill of material; in a manufacturing environment each node on the bill is generally (although not always) a different item with its own replenishment rules such as reorder point, replenishment quantity and lead time etc., but all that MRP requires is that each node along a chain has its own item number (because an item cannot be made from itself in MRP).

Part numbers, or SKUs, can be given prefixes or suffixes to make them location-specific. So an item held, for example, at three locations in the same geographical area could be called a 1234L1, a 1234L2 and a 1234L3. The bills of material for each of those could call up a 1234A1 from an area distribution center and several area distribution centers (A1, A2, A3, etc) could have bill that call up 1234R1 from a regional center. There is no real limit to the number of nodes in the chain so it can reach all the way back to the providing factories if needed.

Because the logic of DRP is the same as that of MRP, the drivers of demand at the end of the supply chain can be the same as for MRP. They can be reorder points, actual sales demand, forecast demand or a mix of all of those. The choice will depend on the length of the supply chain (in days). In very short supply chains, companies can react to actual sales demand, perhaps even by using kanban techniques whilst, if the supply chain is weeks or even months long, forecasts will be essential. Companies will, however, need to decide on reorder frequency as daily deliveries to every distribution location may not be practical.

A major decision will be choosing whether to allow MRP to ripple through all levels of the bill of material (all nodes in the supply chain) in one hit or to stop and await human verification and possibly supply quantity and date adjustment at some or all points. Blowing through all levels in one run will reduce reaction time so it many be that some companies choose manual intervention initially, until confidence in the overall system builds, and then gradually move to automatic running.

Using MRP to calculate demand at each point along the supply chain opens up possibilities for using more of its functionality. Because each node along the chain has its own SKU/item number, it is possible for stock to have different values as it moves along the supply chain. For an individual company, that might not matter because stock at a point of sale is not generally valued differently to stock at a distribution center unless value has been added; as would happen if packaging was added. However; it does open up the possibility of planning the entire supply chain, encompassing multiple companies. The technology is here today to enable this; although not the trust amongst trading partners.

But there are other possibilities, because MRP recommends and facilitates the raising of Works Orders and these can just as easily be Movement Orders. Works Orders hold pertinent information because, in addition to holding data on SKUs/materials required and their quantities, they hold routing information and, for a Movement Order, that would be the resource necessary to complete the movement. A manufacturing routing would tell the system that, for example, to make 1000 items would take 10 hours on machine 'A' but a movement routing could say that, to move 1000 items from the regional distribution warehouse to local outlet 'B' would take 10 pallet spaces on a truck.

That enables two things: it enables the cost of the movement to be estimated (in some cases it can only be an estimate because some movements may not be complete truckloads and, additionally, actual charges may vary if using different hauliers). And, of course, actual charges can be recorded to continually fine-tune those estimates. Routings, because they specify the type and quantity of resources required to carry out a task, can be used for both transport capacity planning and scheduling. Capacity planning will allow companies to calculate how many vehicle journeys and vehicle mikes are required and, if the delivery resource is limited, also allows them to optimize and prioritize those journeys.

So, although specialist SCM systems are now available for some companies there may be considerable value in considering mainstream ERP systems for the task; particularly if they have one already.