Dynamic Machine Routing

 Machine routing identifies the machines that are needed for each of the operation of a given job. Within the JSP or Job Shop Scheduling Problem model, these precedence relations are submitted as technological constraints and they are usually in use as fixed and predetermined. Conversely, owing to the programmable and flexible nature of the classical equipment utilized in computer-controlled systems, machine routing frequently taken a dynamic decision process.

Identify a feature on a part which may be machined on either of two various machines. During forming a routing decision, one may prefer to route the element based upon the stake of the machines at that time. For illustration, specified two machines, r and s, suppose a part can be processed quicker on r than on s; conversely at a specified time, this may be desirable to process this on s since r is busy processing other jobs.

Optional machine routings can be presented by using OR-graphs. Within the OR-graphs G (N, A), each node in N is related along with a processing operation, and the arcs in A are related along with the technological precedence relations among them. A significant characteristic of OR-graphs is such only one outgoing arc should be contained in some solution.  Thus,  at  each  node  in  the  graph,  the  related  decision  problem  is estimating that alternative (outgoing arc) to choose. Notice that, in the traditional JSP, G (N, A) will contain an easy path of operations and no decisions concerning optional plans require considerations. This is well established that by using optional plans specified machine breakdowns or other system state information offers immense results. Via incorporating results optional process plans throughout the schedule generation, one must expect improved results than attempting to resolve process scheduling and planning independently. Obviously, the computational burden will rise. This is due to the additional disjunctions related the technological constraints, additionally to the disjunctive constraints related along with machine capacity, deadlock avoidance and part contact.

This is worthwhile noting that it is rather possible to contain optional process plans (that is processing steps) to be executed on the similar machine. This additional difficulty may also be modeled via extending the OR-graphs presentation mentioned above.