Answers are grounded in this paper and cite their sources.
This paper presents an autonomous robot solution for industrial pipe inspection and contrasts it with traditional human inspection, highlighting advantages in safety, speed, automation, and wireless control. The system uses embedded modules for speech processing, motor control, and real-time data visualization. Key sensor inputs include audio, gas, humidity, temperature, and obstacle distance. A Long Short-Term Memory (LSTM) neural network, trained on Mel-Frequency Cepstral Coefficient (MFCC) features and deployed via TensorFlow Lite, enables real-time, on-device inference. The system achieves high accuracy and responsiveness, making it suitable for real-world industrial scenarios.
No takes yet. Share an insight, caveat, or question.
Kavazović et al. (2025) studied this question.