Demonstrating the Continual Learning Capabilities and Practical Application of Discrete-Time Active Inference