AI optimisation tool for container packing

AI optimisation tool for container packing

Efficient optimisation solver that can pack items into fixed-size containers while minimizing a cost function, e.g. the number of bins required, the surface of the packing etc, whilst adhering to constraints such as preventing items from overlapping, gravity and the need for a physical support, limited rotations, packing stability and weight distribution amongst others. The tool outperforms and scales better than up-to-date optimisation solvers and delivers stable performance even as the problem size increases. The solution uses reinforcement learning to tackle this problem, which is a relatively new method to address problems in this domain.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.

CONTACT COMPANY

SOLUTION FEATURES

  • Best in class performance
    Outperforms and scales better than an up-to-date optimization solver
    Outperform a sophisticated heuristic, i.e. legos
    Stable performance as the problem size increases - strong generalization
  • Efficiency and Scalability
    Efficient Optimization Solver
    Knowledge transfer to larger problem instances
    Approach applicable to many NP-Hard problems

RESOURCES

CATEGORIES

Worldwide Intel® Xeon Scalable CSP - Amazon Web Services CSP - Google Cloud On-premise (Private Cloud, Other) TensorFlow Cross-Industry Models can be trained - data input only required AI Software/SaaS Linux Intel® Math Kernel Library for Deep Neural Networks (Intel MKL-DNN) Proprietary Factory Automation Other Deep Learning