Acoustic Anomaly Detection (AAD) serves as a non-contact approach for machine condition monitoring, yet much of the existing research is evaluated offline and rarely addresses embedded hardware constraints or microphone variations under operational drift. This paper presents a comparative evaluation of on-edge continual AAD across two embedded platforms the ESP32-S3 (with an external I2S digital microphone) and the Arduino Nicla Vision (with an onboard Pulse Density Modulation (PDM) microphone) using a shared detection model. The model utilizes a lightweight ShallowNet encoder and contrastive prototype learning to map spectrograms into a 32-dimensional embedding space. Both boards run an identical TensorFlow Lite for Microcontrollers (TFLM) encoder, performing on-device prototype learning updated via an exponential moving average. A cloud pipeline orchestrated through Firebase and MQTT enables remote data collection and retraining. Evaluated via a controlled speaker-replay protocol using standardized Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge Task 2 machine sounds, the ESP32-S3 achieved 82.35% to 95.00% accuracy across five machine types, while the Nicla Vision achieved 79.41% to 95.00%. The two platforms demonstrated comparable performance, confirming that the lightweight prototypical architecture transfers reliably across distinct hardware and sensor tiers.
Figures & tables
Parameter
Symbol / Identifier
Value
Audio Sample Rate
SAMPLE_RATE
16,000 Hz
FFT Frame Length
FRAME_LEN
480 (30 ms)
FFT Frame Hop
FRAME_STEP
320 (20 ms)
Number of Temporal Frames
NUM_FRAMES
49
FFT Size
FFT_SIZE
512
Frequency Buckets
NUM_BUCKETS
43
TABLE I: Shared System Hyperparameters and Configuration
Fig. 1: Encased Arduino Nicla Vision deployment platform featuring external antenna connectivity and secure acrylic mounting for physical playback evaluation.
Fig. 2: Controlled physical speaker-replay experimental pipeline. Calibrated DCASE benchmark machine recordings are emitted via a reference loudspeaker in a closed room and captured by edge microcontrollers running local TFLM inference.
Machine Type
Benchmark Source Dataset
Normal ( n )
Abnormal ( n )
ToyCar
ToyADMOS2 [ 3 ]
17
17
Toy Train
ToyADMOS2 [ 3 ]
17
17
Fan
MIMII / MIMII-DG [ 1 , 5 ]
17
17
Gearbox
MIMII-DUE / MIMII-DG [ 4 , 5 ]
20
20
Slider
MIMII / MIMII-DG [ 1 , 5 ]
20
20
Total
DCASE Task 2 Suite
91
91
TABLE II: Characteristics of the Evaluated DCASE Machine Datasets
Information Sciences, CAI-3, Los Alamos National Laboratory, Los Alamos, USA · Information Sciences (CAI-3), Los Alamos National Laboratory, Los Alamos, NM, USA · AeroVironment Inc., Albuquerque, USA +7
computer science, Deggendorf Institute of Technology, Deggendorf, Germany · Computer Science and Engineering, University of West Bohemia, Pilsen, Czech Republic · The Applied College, King Abdulaziz University, Jeddah, Saudi Arabia
School of Software and Microelectronics, Peking University, Beijing, China · Independent Researcher · Collaborative Innovation Center for Western Ecological Safety, Lanzhou University, Lanzhou, China +2