Ubiquitous embedded systems such as wireless sensors and IoT devices must process significant amounts of digital data under strict hardware constraints. Ensuring secure access to this data is critical, especially in such resource-limited environments. Therefore, cryptographic schemes must be designed to operate efficiently within these constraints. However, traditional public-key cryptography algorithms are increasingly vulnerable to quantum computing attacks. In response, post-quantum cryptographic schemes have been developed to withstand threats posed by quantum computers. This introduces a new challenge: developing lightweight PQC schemes that can be deployed on constrained embedded platforms. In this study, we present newly optimized implementations of the SMAUG-T key encapsulation mechanism (KEM) at three NIST security levels, SMAUG-T 1, SMAUG-T 3, and SMAUG-T 5, to evaluate its suitability for lightweight embedded systems. Although the original paper of SMAUG-T targets AVX2-based high-performance processors, their applicability to embedded systems has not yet been studied in detail. To address this gap, we implement SMAUG-T on an ARM Cortex-M4 microcontroller and evaluate its performance in accordance with criteria from related studies. Our results show that SMAUG-T has strong potential for efficient deployment on resource-constrained embedded devices.