A lot of manufacturing companies today still use a whiteboard or spreadsheet of some kind to manage their production scheduling. Scheduling managers for those companies know their products. They know the processes which go into those products. They know the production standards necessary to convert certain raw materials into those products using those processes. And they know the people who run those processes and what they’re capable of.

In most situations however, managing production scheduling using this method can be considerably overwhelming for someone to take care of on their own. Downtime occurs on machines intermittently across the production floor. Schedules subsequently slip. Or a rush order comes in and everything currently scheduled on various machines or within a given department will need to be pushed back. It gets to the point where it’s a full-time job, if it isn’t one already.

Automating the production scheduling of a manufacturing facility will likely become a necessity, especially if a facility or company grows as time goes on. Smaller facilities would do well to start with even a “manual” software-based production scheduling system. As with most of these systems, jobs are laid out as tags on a virtual “planning board”. Workcenters are listed down the left side of this planning board, and a timeline is laid out across the top. The tags for the individual jobs then list the job number, quantities, the customer, individual or master part numbers … just about anything needed for the individual using the planning board.

Behind the scenes of a production scheduling system is where a lot of other work takes place. Processes can be scheduled as part of a larger work order, personnel can be scheduled individually, holidays, lunch periods and shift changes can be accounted for, etc. Other calculations can also take place, such as the longer times necessary to bend or roll a thicker piece of steel, ensuring jobs don’t overlap where they shouldn’t.

A somewhat more automatic production scheduling system, using a certain level of artificial intelligence, can take a look at the current schedule and figure out the next logical places to run a given set of processes. It will break an order down into its individual processes, look at which machines are capable of running which process with which material, figure out the earliest a process can run and on which machine, and schedule everything as needed with minimal intervention from the production scheduler.

A production scheduling system which uses artificial intelligence to perform the bulk of the scheduling will be unique to a given facility. Using constraints based on fixture availability, the number of cells in a given workstation, or even the groupings of colors of the items being manufactured, the artificial intelligence will shuffle workorders in a large “shell game” looking for the schedule which gives the best results. Once this schedule is found, which may mean the AI has looked at tens of thousands of possible schedules for those workorders, the operator can then settle on that schedule and implement it.