This chapter introduces a brain-inspired sequence learning model designed to handle the temporal nature of real-world cognition, where patterns unfold over time. Tasks such as interpreting language or responding to dynamic environments require the ability to process sequences, not just static representations. The proposed model adopts a lightweight, biologically inspired approach based on Sparse Distributed Representations (SDRs), aiming to emulate the interaction between short-term and long-term memory through mechanisms such as association, forgetting, and prediction. The model operates without the need for complex training or large datasets, making it suitable for low-power and real-time applications. Its design incorporates hash-based hierarchical SDR encoding and logical bitwise operations, enabling fault tolerance and pattern recognition under noisy conditions. To validate the concept, the model is implemented in a hardware simulation using low-cost components, demonstrating its feasibility for embedded and edge computing environments. Compared to conventional deep learning methods, this approach offers advantages in interpretability, online learning, and resource efficiency. The results highlight the potential of brain-inspired memory models for sequence processing tasks in constrained settings, offering a practical foundation for future applications in natural language processing and beyond.