Automatic Speech Recognition (ASR) systems are becoming more commonplace in real-world applications. Despite this increase in usage, their robustness in noisy environments remains problematic for correct word identification. This study offers an automated program to test ASR systems alongside different background noise. It tests several ASR systems (Whisper-Small, Whisper-Medium, Whisper-Large-V3-Turbo, Parakeet 0.6b, Canary 1b, and Commonvoice-Wav2Vec-EN) across five total noise conditions (white noise, speech shaped noise, and three industrial noises) at varying levels of loudness (64–79 dB). Results indicate that ASR systems have significantly reduced word recognition across all noise levels with industrial machine noise poising a greater challenge than other types of noise at moderate intensities. Additionally, opting to avoid enhancements to ASR improved performance overall, particularly for female speech.