The secret of world-record-breaking optimization technology

Here’s a glimpse of five of the optimization technologies that form a significant part of Quintiq’s optimizer.

The first is one you might be encountering for the first time –
  • Quill
    Quill (Quintiq Intelligent Logic Language) is Quintiq’s proprietary configuration language. Quill is at the centre of Quintiq’s optimizer. It’s the glue that holds the optimization technologies together, allowing them to work as an effective, efficient whole.

    Key strengths & capabilities:
    • Allows easy expression of construction heuristics and local search heuristics – this matches closely the processes that human planners follow when they build up solutions from scratch and when they locally improve solutions
    • Combines optimization technologies in one single optimizer – to produce solutions that draw upon the strengths of all five elements
  • Mathematical Programming
    Mathematical programming is mature and stable. It solves puzzles by choosing the right set of building blocks that together constitute the best solution, even where there are complex constraints and relationships. For example, it's used in Quintiq's logistics solution to determine the best possible combination of routes to ensure the shortest distance and highest service level.
  • Constraint Programming
    Constraint programming is particularly effective in dealing with optimization puzzles when finding any solution is already difficult. It works by continuously eliminating potential solutions according to the constraints. Constraints can be non-linear. Constraint programming is particularly useful in solving scheduling puzzles.
  • Path Optimization Algorithm
    Path optimization algorithm, known to Quintiq developers as POA, is a proprietary large-neighborhood-search (LNS) technology. LNS explores much larger neighborhoods than local search and as such is less likely to get stuck in a locally optimal solution that is far away from the globally optimal solution. POA is primarily used for vehicle routing and manufacturing scheduling.
  • Graph Programming
    Many practical puzzles have a graph component and accordingly we utilize graph programming to find solutions. Graph programming is used to tackle resource-constrained shortest-path problems encountered in puzzles such as vehicle routing and crew diagramming.

For Quintiq's latest world records in optimization, click here