Exception handling
View Patent ↗A system that re-plans jobs based at least in part on user preferences in response to system component errors includes an exception handler that receives an exception from one of a plurality of components executing a plan to process a job. The system further includes a planner that creates a new plan for the job based at least in part on a model-based planning technique and at least one user preference.
1. A system that re-plans jobs based at least in part on user preferences in response to system component errors, comprising:
an exception handler that receives an exception from one of a plurality of components executing a plan to process a job; and
a planner that creates a new plan for the job based at least in part on a model-based planning technique and at least one user preference and further creates new goals for currently executing portions of the plan operating in real-time, and wherein the new plan is forward simulated for an amount of time that depends on a number of plans that need to be re-built, the planner determining the new plan starting from a new location, of the system, for the new goals using a chained best first search algorithm, and wherein the new plan is sent to an execution engine to replace the plan which was the previous plan.
2. The system as set forth in claim 1 , wherein the new plans are based on the new goals and include one of discarding all partially processed product; discarding all partially processed product corresponding to the component that threw the exception; discarding partially processed product that violate job integrity; and only discarding partially processed product that needs to be reproduced.
3. The system as set forth in claim 1 , wherein the new plan is created on-line while the system continues to run.
4. The system as set forth in claim 1 , wherein the new plan is further based on at least two of trade-offs between saving resources and exception handling time/effort, a state of the system, a cost model, the user's preferences, and the types of exceptions.
5. The system as set forth in claim 1 , wherein the planner further creates one or more plans in order to process one or more jobs based on one or more capabilities of the plurality of components and one or more constraints on the one or more capabilities.
6. The system as set forth in claim 1 , wherein the planner further handles batches of jobs that follow a particular order at substantially the same time.
7. The system as set forth in claim 1 , wherein the new plan includes determining the new goals for currently executing portions of the plan according to real-time constraints and relations with a malfunctioning portion of the plan.
8. The system as set forth in claim 1 , wherein the planner switches to a re-planning mode to create the new plan and returns to a normal mode to execute the new plan.
9. The system as set forth in claim 1 , wherein the system is one of a printing platform, a manufacturing system, a packaging system, a logistic system, a routing system, a process, and an analysis system.
10. The system of claim 1 , wherein the exception is one of a plurality of possible exceptions and use of the model-based planning technique and the at least one user preference mitigates a need to determine rules for all possible exceptions.
11. The system of claim 1 , wherein the planner that creates the new plan uses an algorithm that corresponds to the at least one user preference.
12. The system of claim 11 wherein when the at least one user preference is to increase a speed at which a job is processed, the algorithm used by the planner is a time optimizing algorithm, and when the at least one user preference is to lower the cost of performing the job, the algorithm used by the planner is a cost optimizing algorithm.
13. A method for creating new plans for jobs based at least in part on user preferences in response to exceptions, comprising:
receiving an exception associated with a component implementing a plan;
cancelling all unlaunched plans;
creating new goals for currently executing portions of the plan operating in real-time that threw the exception, based at least in part on a user preference provided by a user;
forward simulating currently executing plans;
creating new plans for the new goals starting at the forward simulation time, wherein the new plan is forward simulated for an amount of time that depends on a number of plans that need to be re-built, the planner determining the new plan starting from a new location, of the system, for the new goals using a chained best first search algorithm, and wherein the new plan is sent to an execution engine to replace the plan which was the previous plan.
14. The method as set forth in claim 13 , further including anticipating an affect the forward simulation will have on future failures.
15. The method as set forth in claim 13 , further including creating the new goals based on one of the following approaches: purging executing portions of executing plans; only purging executing portions of plans associated with the exception; and only purging executing portions that are rebuilt.
16. The method as set forth in claim 15 , further including selecting the approach based on at least one of a user preference, a goal, a system model, a type of exception, a cost, a benefit, an inference, a probability, a classifier.
17. The method as set forth in claim 13 , further including:
estimating a re-planning time based on a number of new plans that will be created and a time it takes to create each new plan; and
basing the forward simulation on the estimated re-planning time.
18. The method as set forth in claim 13 , further including executing the new plan to process the job.
19. A xerographic system that re-plans jobs based at least in part on user preferences in response to exceptions, comprising:
an exception handler, of the xerographic system, that receives an exception from one of a plurality of components executing a plan to process a job;
a planner, of the xerographic system, that creates a new plan for the job based at least in part on a model-based planning technique and at least one user preference provided by a user and further creates new goals for currently executing portions of the plan operating in real-time, and wherein the new plan is forward simulated for an amount of time that depends on a number of plans that need to be re-built, the planner determining the new plan starting from a new location, of the system, for the new goals using a chained best first search algorithm; and
an execution engine, of the xerographic system, that executes the new plan, wherein the new plan is sent to the execution engine to replace the plan which was the previous plan.
20. The xerographic system as set forth in claim 19 , wherein the plan is forward simulated for an amount of time that depends on a number of plans that need to be re-built and the new plan is created based on the forward simulated time.
21. The xerographic system as set forth in claim 19 , wherein the planner anticipates an affect the forward simulation to mitigate potential future exceptions.