Lesson

A2 Validate First: Design a Useful Planning Comparison

On this page

One machine, two orders and a maintenance window. This small scenario illustrates why a plan comparison starts by explaining how conditions have changed, before interpreting a system's result.

This lesson develops three judgments: whether current conditions permit on-time completion, what a change affects, and which data, configuration or management decision could address it.

The exercise uses teaching data, with processing time and available capacity calculated by hand from the stated conditions. The second half uses existing recorded cases to explain how to conduct a similar comparison in DecisioWorks.

1. State the Decision

Two orders must be completed today on work center WC-1. Each product has one operation and can only use WC-1. The machine processes one batch at a time. All required materials and other resources are available. Both orders are due by the end of today.

The exercise has only one planning period: today. Production cannot be moved to an earlier period. Each product forms one complete batch that runs continuously. No initial setup is needed. Changing from A to B or B to A takes 40 minutes and occupies WC-1. There are no other changeovers.

The baseline uses today's normal working time, the stated quantities and the specified process. Overtime, alternative equipment, batch splitting and later delivery require separate decisions and have not been included.

2. Input Data

Orders and Processing Conditions

Order	Product	Quantity (units)	Processing time (minutes/unit)	Work center	Due
O-1	A	20	10	WC-1	End of today
O-2	B	20	11	WC-1	End of today

Calendar and Change

Condition	Baseline	With maintenance
Gross shift time before maintenance (minutes)	480	480
Planned maintenance (minutes)	0	30
Net available time (minutes)	480	450
One changeover between products (minutes)	40	40

Maintenance occurs before production starts, leaving 450 continuous minutes. If it occurred in the middle of production, the exercise would also need to check whether continuous batches fit around the maintenance window. Total available time alone would answer only part of that question.

These business tables explain the calculation. When preparing data.db, also provide the object identifiers, time units and process-resource relationships required by the standard interface.

3. Separate Facts, Constraints and Possible Measures

Category	In this exercise	Meaning
Business facts	Order quantities, processing times and maintenance period	Conditions that the data must represent accurately
Required conditions	All quantities completed today, no overlapping use of the machine, continuous batches	The basis for assessing whether current commitments can be met
Preferences to compare	Sequence choices or other supported objectives among feasible plans	Trade-offs to examine when feasible alternatives exist
Measures requiring approval	Overtime, later delivery, alternative resources or batch changes	Options whose practical feasibility must be confirmed before updating data and configuration

Hard constraints prohibit violations. Soft constraints allow deviations at a defined cost. If capacity is modeled as a soft constraint and a candidate plan is returned, inspect whether it includes overload and how that overload would be handled in practice.

4. Check the Arithmetic First

Product A requires 20 × 10 = 200 minutes. Product B requires 20 × 11 = 220 minutes. Including one 40-minute changeover, total occupancy is 460 minutes.

Check	Baseline	With maintenance
Processing time (minutes)	420	420
Changeover time (minutes)	40	40
Total occupancy (minutes)	460	460
Net available time (minutes)	480	450
Spare time or shortfall (minutes)	20 spare	10 short

The baseline has a simple feasible sequence: run A for 200 minutes, change over for 40 minutes, then run B for 220 minutes. The total is 460 minutes. After maintenance is added, either sequence still needs 460 minutes, but only 450 continuous minutes are available. Under the stated conditions, both full-quantity and same-day commitments cannot be met together.

This conclusion applies to the simplified conditions of this exercise. Multiple operations, alternative resources, batch splitting and different calendars require additional relationship and time-window checks.

5. Apply the Method in DecisioWorks

Choose the existing stamping planning case or injection-molding scheduling case. Establish a baseline, then select one change supported by that case and its interface. An orchestration recipe specifies the actions, settings and data used for execution and later review.

Step	Action	Record to keep
1. Confirm the question	State the planning period, objects and condition to compare	Scenario brief
2. Inspect the inputs	Check units, processes, capacity and material relationships in a copy of the sample	Data version, relevant tables and checks
3. Confirm configuration	Record enabled constraints, objectives, computing resources and allowed run time	Baseline configuration
4. Run the baseline	Use the page or recipe entry point supported by the case	Status, action history, result tables and analysis
5. Change one condition	Modify a supported objective parameter, or change a business condition in a separate data copy	The change, its reason and the revised version
6. Compare and explain	Check status and constraints before business measures and detailed rows	Conclusions, decisions requiring confirmation and the next step

In the Scenario Lab, use the configuration exposed by the page to run a baseline and adjust parameters. For calendar or process-data changes, work on a separate data copy and use the integration and execution entry point supported by that scenario. Keep parameter experiments separate from changes to business data.

For an objective-parameter comparison, keep the input data unchanged. For a data change such as maintenance, preserve the original copy and keep other parameters consistent where possible. Keep thread counts, time limits and other execution conditions consistent; repeat runs where appropriate to assess stability.

6. Read the Result in Order

Status: was a usable plan produced? If not, was infeasibility proved, did time run out, or was there an input, environment or authorization issue?
Conditions: which constraints were active? Were overload, lateness or unmet demand allowed?
Business measures: how much was produced or left short, when was work completed, and how much resource time was used?
Objectives and details: how did weights influence the trade-off, and which orders or work centers changed?
Decision: can the plan be executed, and who must confirm any measures it requires?

Objective totals calculated with different weights use different evaluation scales. Compare actual business measures first. If comparing total scores is useful, evaluate both results with the same objective definitions, directions and weights.

7. Exercises and Suggested Answers

Exercise	Suggested answer
Will increasing the delivery-objective weight tenfold remove the 10-minute shortfall?	A weight expresses preference; it does not add time. With full quantity and delivery enforced as hard conditions, the shortfall remains. If deviations are allowed, explain where the unmet requirement moves
Will producing B before A solve the problem?	Both changeover directions take 40 minutes here, so total occupancy stays the same. Recalculate if directional changeover times differ in another case
After 10 extra minutes are approved, what still needs confirmation?	Here, confirm that the extra time extends the existing 450-minute window continuously and that staffing and materials remain available. The resulting 460 minutes accommodate the stated sequence. Other cases also need checks on their operations and resources
If a program finds no plan in 60 seconds, can the result be described as a business infeasibility?	Inspect its status and proof information. Search time exhaustion and proven infeasibility must be described separately
If all quantities are produced after the change, can that be described as improved delivery?	Completion times must also be compared with due dates. Measure quantity fulfillment and on-time delivery separately

8. Carry the Learning into Implementation

This exercise leaves three reusable assets: a way to describe production conditions in data, a fair comparison method, and a process for deciding what to do with the result.

As the scope expands to more orders or resources, the same data checks and comparison method can be used again. During regular operation, changes to maintenance, objectives or rules can build on the recorded work. DecisioWorks supplies data, configuration, execution and analysis capabilities; manufacturers and partners connect them to their own processes and confirm the operational decisions.