Items are the physical (and non-physical) entities that flow through the production system. Depending on the production system type, the items will vary. An engineering production system may have items in the form of drawings, specifications, and approvals. A fabrication or assembly production system may have items in the form of raw materials, parts and assemblies. A supply chain production system may gave items in the form of shipments, bundles, and containers components being transported. Construction production systems may have items such as foundations, steel structures, permanent equipment, and commissioned systems.
To Create an Item:
- Open the Data Entry pane on the left side of the User Interface
- Select Item
- Select the + button and fill in the input fields
- Select Create
To Edit an Existing Item:
- Double click the item or select the item and then select the pen button
- Edit the input fields and select Close
To Delete an Existing Item:
- Select the item and then select the X button
Item Data Element Description
| Data Element | Description | Constraints |
|---|---|---|
| Description | Description of the item | |
| ID | Unique ID for item | Unique to all other items |
| Unit of Measure | The ID for the unit of measure, selected from the Unit of Measure table | UOM table |
| Fixed Order Policy | If this is checked, current reorder points and reorder policies will not be changed as part of the inventory optimization calculation, i.e. existing policies are frozen. | Check mark |
| Purchased Product | Indicates if the item is purchased from a 3rd party and not manufactured at the plant. This determines if the Production Optimizer should look for a routing when modeling capacity. If this is checked, Bill of Material explosion will not attempt to explode demand past this item. | Check mark |
| Transfer Batch (Units) | The number of units that is moved or transferred between steps in the routing. If this changes throughout the process, choose the largest value. | Integer ≥ 0 |
| Raw Unit Cost ($) | The total raw material cost of one unit | ≥ 0 |
| Total Unit Cost ($) | The cost of one unit, including labor and raw materials | ≥ 0 |
| On Hand (Units) | The current number of units (UOM) on hand. Used to compare against CONWIP Level if Bottleneck CONWIP protocol is set. | Integer ≥ 0 |
| Historical Cycle Time (Days) | Average historical cycle time in days | ≥ 0 |
| Revenue Schedule | Weekly revenue schedule used as an input to Value Model analysis | Revenue table |
| Average Number of Orders (Per Period) | This is the number of orders (in units) released during the schedule period. It is also equal to the number of orders started. If there is no information on average order size, set this field equal to demand in units during the schedule period and set Average Order Size to 1. | > 0 |
| Variance of Orders (Per Period) | Variance of orders or mean squared error of forecast. If no information on Average Order Size, this becomes the variance of demand. | > 0 |
| Average Order Size | The average size of the orders released. This is applicable for bulk demand situations where partial shipments are acceptable. | Integer > 0 |
| Standard Deviation of Order Size | The standard deviation of the order size. Applicable to bulk demand situations. Enter 0 when dealing with single demands. | > 0 |
| Average Lead Time (Days) | The amount of time required to receive a stock order into stock after the inventory position for the stock point reaches the reorder point. In other words, it is the time between the order is placed and the time is arrives to the stock point | > 0 |
| Standard Deviation of Lead Time | Standard deviation of the lead time. | ≥ 0 |
| Include Backorder Time | If checked the system will include backorder time in the replenishment time calculation If a BOM has been created and this box is checked, the replenishment time for each item will include the cycle times of the items beneath them in the BOM hierarchy | Check mark |
| ROP Method | Flag to indicate how the order time values are computed using the entered values Fixed: Safety Stock is calculated using the entered Reorder Point (ROP) value Time-Phased: ROP is calculated using the entered Safety Stock value, average daily demand, and lead time | Fixed, Time-Phased |
| Safety Stock Quantity (Units) | Number of units kept on-hand to buffer variability in demand | Integer ≥ 0 |
| Current Reorder Point (Units) | Inventory Position at which the next order is triggered | Integer ≥ 0 |
| Lot Size Method | Flag to indicate how the lot size values are computed using the entered values Days of Supply: Reorder Quantity (ROQ) is calculated based on the entered Days of Supply Value Fixed Order Size: Days of Supply is calculated based on the entered ROQ value. | Days of Supply, Fixed Order Size |
| Days of Supply (Days) | The number of days that MRP (material requirements planning will look out past the first occurrence of a negative projected on hand balance. MRP will accumulate all demand in the days specified and use that as the order amount. | > 0 |
| Current Reorder Quantity (Units) | Standard order quantity of the item | Integer > 0 |
| Minimum Reorder Quantity (Units) | The minimum order quantity allowed for this item | Integer ≥ 0 |
| Maximum Reorder Quantity (Units) | The maximum order quantity allowed for this item | 0 – 999,999 |
| Reorder Quantity Increment (Units) | The increment order quantity allowed for this item | Integer > 0 |
| Minimum Fill Rate (%) | A user-specified lower limit on fill rate. When inventory optimization is done, it will ensure that the item’s fill rate will be no less than specified here. | < 100 |
| Planned Lead Time (Days) | The time planned between a new order for a part and the furnishing of that part. | > 0 |
| Order Cost ($) | Out of pocket cost associated with orders. | ≥ 0 |
| Made to Stock | If checked, the inventory optimizer will include this item in the stock point optimization | Check mark |
Production System Demand
Calculating demand is a necessity for modeling any production system. This section describes how to calculate demand for Project Production Systems and Manufacturing / Supply Chain Production Systems
Demand is the consumption (in items per period) “seen” by the items in the production system. Demand for each item is given in terms of a mean and a variance. There are a few topics that come up repeatedly in discussing item demand and they will be addressed here to provide guidance for the user.
Demand for Projects
Demand in a project environment is almost always deterministic. As Front-End Engineering and Design (FEED) concludes, and Detailed Engineering kicks off, rough quantities for demand planning are already decided and will become available to the planner. Accurate demand quantities will be available after material takeoffs are accepted.
Once the project milestones and schedule are determined, a timeline is in place. This timeline and the quantities per time can be used to calculate the rate of demand for each element in a project production system. Thus, the rates of demand over the course of engineering, fabrication, construction, and commissioning are set by 2 factors:
- Total quantity of production units – these are things such as engineering documents, steel beams, pipe spools, electrical cable, modules, hydrotest packs, and other components of the project.
- Start and Finish dates – these dates are typically taken from the project schedule and are driven by project objectives and constraints. These set the speed at which the production units must be produced.
To calculate demand for a project, divide quantity by time. Demand can be a uniform value over the time horizon, or it may follow a different distribution.
To calculate demand, divide the number of production units that need to be produced by the time between the start and finish dates. For example, an engineering team needs to produce 1500 isometric drawings between July 1st and October 31st. This represents an average of 17.5 working weeks, the assuming engineering team works Monday through Friday. Demand for this engineering team is 85.7 drawings per week (1500 drawings / 17.5 weeks). This can be rounded up to 86 for the sake of working with whole numbers.
A demand of 86 drawings per week represents a uniform demand over the 17.5 weeks. In the real world, demand profiles can take on many shapes and may not be uniform. If not uniform, project teams should investigate if smoothing demand (making more uniform) is possible. This will reduce variability in the production system and make capacity planning easier. If smoothing demand is not possible due to its impact on other project production systems and if access to additional labor and equipment takes time, project teams should plan for peak demand periods and accept low utilization during low demand periods.
Demand for Manufacturing and Supply Chain
There are two major sources of information about demand: history and forecast. Calculating demand and variance from historical data is straightforward. Generally, monthly demand is used. An example of a monthly demand calculation is provided in Table 1. Weekly demand is better to use if it is available because there is no need to compensate for the different number of days in each month. This applies particularly for companies that use a 4-4-5 calendar.
| Jan | Feb | Mar | Apr | May | Jun | Total | Mean | Var. | |
|---|---|---|---|---|---|---|---|---|---|
| Item 1 | 100 | 200 | 500 | 60 | 160 | 300 | 1,320 | 220 | 25,760 |
Forecast Error
Past performance is not always a good predictor of future performance. A common question is, “How will the Production Optimizer address future demand?” The answer is: with a forecast. However, there are two things to keep in mind about forecasts:
- The forecast is always wrong
- The forecast will always change
Companies can use forecast error for demand variance. Forecast error should be measured using the Mean Squared Error (MSE) of the forecast. Additionally, the horizon over which the forecast error should be measured should be equal to the replenishment time of the item. If an item has a replenishment time of 4 weeks, the MSE should be measured for the four-week forecast for that item.
If the MSE is consistently greater than the historical variance of demand, there is more variability being introduced into the inventory control process with forecasting than would be by simply using historical variance. In other words, in this case, forecasting makes things worse.
Variance to Mean Ratio
When it comes to forecasting demand variance, historical variance is a very good indicator of future variance. In that case, using the variance-to-mean ratio provides a very good estimator of future variance. The application is simple. Calculate historical mean demand and variance of demand, or X-bar and S2. Estimate future mean demand from the forecast, μF. Divide historical variance by historical mean and multiply this by the future mean demand to get an estimate of future variance of demand:
σ2F, ≈ μF * S2 / X-bar