IP Technology - Adaptive Optimization®

Quantum Leap Adaptive Optimization® is a patented general-purpose problem-solving system that is especially suited to:

1. • Unusually complex, non-traditional optimization problems

2. • Situations based on changing or unstable requirements

3. • Data-driven scenarios

Quantum Leap Adaptive Optimization has been applied in a wide range of manufacturing and military applications, including tradeoffs when designing new vehicle concepts, planning, allocation, and scheduling of assembly lines and manufacturing capacity.  The system does not require its users to be optimization experts; users can focus on what they are trying to optimize.

1. • Adaptive Optimization does not require users to purposefully select an optimization technique or solution in advance (separates problem statement from solution strategy).

2. • Adaptive Optimization doesn’t believe that users should jump through hoops to describe a problem so it can be processed by a pre-determined technique. (Solving technologies often constrain the way we view problems, and inhibit us from modeling the real world)
   

3. •   Adaptive Optimization focuses on enabling human-oriented descriptions of hard problems (not optimization framework-oriented descriptions). This drastically simplifies updating a problem because of a change since a human doesn’t have to relearn a problem encoded for computer use.

More About Adaptive Optimization

Quantum Leap Adaptive Optimization automatically combines and uses 30 different techniques to solve optimization problems.  Most optimization problems today are solved by pre=seeding one technique, usually linear programming (LP) or mixed integer programming (MIP).  However, most real-world problems are more complex than that. Therefore problems usually get simplified to fit one common solving technique - resulting in solving the wrong business problem!

Furthermore, as business problems are translated to mathematical representations, the translation process can result in producing many more math variables in order for the common solving technique to be applied -- making the math problem even more complex and difficult to maintain.  Quantum Leap advises customers to represent their business problems accurately, and model them separately from the solving technique.  Then use the best solving technique that matches the problem.  This results in a system that is easier to maintain and update. Quantum Leap Adaptive Optimization supports any optimization analytics required by the business.

 Quantum Leap Adaptive Optimization is transparent and automatic.  The solving techniques used are automatically selected, based on the type of problem.  Automatic transparent selection means that users don’t have to be math wizards to understand which optimization technique is the best to solve a particular problem. 

Quantum Leap Adaptive Optimization® automatically combines 30 techniques to solve any optimization problem

Although capable of handling both simple and complex optimization problems, Quantum Leap Adaptive Optimization works best in problems that are unusually complex and data driven.  Adaptive Optimization fits well in applications such as:

1. •Real-time monitoring and planning of manufacturing supply chains
   

2. •Strategic planning "what if" business tool for resource allocation
   

3. •General-purpose server "bolt-on" analytic engine - wide variety of solving techniques is required
   

4. •Portfolio construction, optimization, and balancing
   

5. •Maintenance Planning
   

How does it work?

Quantum Leap Adaptive Optimization works by automatically selecting an optimization technique and continually refining the optimum point with additional techniques.  This patented approach is a unique optimization system, since most other optimizers allow the user to pre-select one of the techniques from a library.  Most optimization solvers are designed for solving unconstrained problems, Quantum Leap Adaptive Optimization is unique in that it handles constraints plus optimization.

Quantum Leap Adaptive Optimization runs on a standard PC system, and also exploits parallel processing architectures to easily scale up if horsepower is available.  Quantum Leap Adaptive Optimization is written in Java and supports most common Windows and Linux operating systems.

Quantum Leap Adaptive Optimization is protected by U.S. Patent Numbers 1,990,804, 5,195,172 and 5,428,712