0 Replies Latest reply on Mar 24, 2011 1:14 PM by Janice Kakazu

    Constraint-based planning – can you avoid creating a ‘black box’?

    Janice Kakazu Expert

      Constraint-based planning is very seductive.  It sounds great – let the system do all your planning for you!  It will take into account all your constraints, only load up to the max constraint allowed, and then reschedule orders either forward or backward or any which way you desire, to come up with the perfect plan!  The only problem is – will you understand and trust the results when it’s all done? 

      I have seen other instances where complex planning algorithms become a ‘black box’ for the average planner.  They can be so complex that the planner who is trying to understand and make use of the results can’t figure it out.  While the planning results may be ‘right’ per the logic and algorithm, the results may be counter to the planner’s intuitive sense of what it should be, and we all know what happens when a user doesn’t believe the answer!  The ‘system’ loses credibility and the users won’t follow its recommendations.  In addition, due to some data issue, the results may actually be ‘wrong’, but it is very difficult to figure out what is causing the error. 

      However, it is undoubtedly true that by taking into account constraints in planning, a company can save money and operate more efficiently with better customer service.  Given the risks and rewards, is it worth implementing constraint-based planning?  Is there a way to make it more transparent and less confusing/complex for the users?  What do you think?