Supply chain and logistics systems are very complex, especially when they involve asset distribution at national and global scales. Maximising revenue while minimising transportation costs is a multi-objective problem for which no adequate tools exist nowadays. Traditional AI algorithms can create prediction models, but are not sufficient to support decision-making processes by the user.

Skymantics is using Reinforcement Learning, a novel family of AI algorithms, to maximise rewards in a complex and dynamic environment which changes with user decisions. We develop models for different logistics and distribution problems, such as our recent prototype of forestry management model for New Brunswick region (Canada), developed for NRCanada. This type of algorithms provides the user with automated decision support tools for tactical and strategic planning when multiple activity centers, users, selling prices, and transportation costs were involved.