Addressing Power Issues in Biologging: An Audio/Inertial Recorder Case Study

Miquel, Jonathan; Latorre, Laurent; Chamaillé‐Jammes, Simon

doi:10.3390/s22218196

Cited by 2 publications

(4 citation statements)

References 28 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Higher current peaks correspond to SD writing events. In a previous study [4], it was observed that SD writing power efficiency improves when large data chunks are stored at once. Therefore, data are always buffered prior to SD writing, with a constant buffer size of 30 SD sectors (15 kB).…”

Section: Embedded Power Consumptionmentioning

confidence: 97%

“…This MCU is designed around an ARM Cortex-M4 core, featuring a hardware Floating Point Unit (FPU). Further details on hardware choices are provided in [4].…”

Section: Hardware Architecturementioning

confidence: 99%

“…In this context, over the past few years, we designed our own solution for embedded audio recording. Our device engineering is strongly driven by weight over autonomy ratio and exhibits cutting-edge power efficiency, being able to collect about 900 h of 16-bit 8 kHz audio data (~50 GB uncompressed audio samples) from a single small 1800 mAh Li-Ion battery [4]. In this application, recorded data are not streamed over an RF link, but rather stored onto an SD memory card for obvious energy-saving reasons.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Energy-Efficient Audio Processing at the Edge for Biologging Applications

Miquel

Latorre

Chamaillé‐Jammes

2023

JLPEA

Self Cite

View full text Add to dashboard Cite

Biologging refers to the use of animal-borne recording devices to study wildlife behavior. In the case of audio recording, such devices generate large amounts of data over several months, and thus require some level of processing automation for the raw data collected. Academics have widely adopted offline deep-learning-classification algorithms to extract meaningful information from large datasets, mainly using time-frequency signal representations such as spectrograms. Because of the high deployment costs of animal-borne devices, the autonomy/weight ratio remains by far the fundamental concern. Basically, power consumption is addressed using onboard mass storage (no wireless transmission), yet the energy cost associated with data storage activity is far from negligible. In this paper, we evaluate various strategies to reduce the amount of stored data, making the fair assumption that audio will be categorized using a deep-learning classifier at some point of the process. This assumption opens up several scenarios, from straightforward raw audio storage paired with further offline classification on one side, to a fully embedded AI engine on the other side, with embedded audio compression or feature extraction in between. This paper investigates three approaches focusing on data-dimension reduction: (i) traditional inline audio compression, namely ADPCM and MP3, (ii) full deep-learning classification at the edge, and (iii) embedded pre-processing that only computes and stores spectrograms for later offline classification. We characterized each approach in terms of total (sensor + CPU + mass-storage) edge power consumption (i.e., recorder autonomy) and classification accuracy. Our results demonstrate that ADPCM encoding brings 17.6% energy savings compared to the baseline system (i.e., uncompressed raw audio samples). Using such compressed data, a state-of-the-art spectrogram-based classification model still achieves 91.25% accuracy on open speech datasets. Performing inline data-preparation can significantly reduce the amount of stored data allowing for a 19.8% energy saving compared to the baseline system, while still achieving 89% accuracy during classification. These results show that while massive data reduction can be achieved through the use of inline computation of spectrograms, it translates to little benefit on device autonomy when compared to ADPCM encoding, with the added downside of losing original audio information.

show abstract

Section: Embedded Power Consumptionmentioning

confidence: 97%

“…This MCU is designed around an ARM Cortex-M4 core, featuring a hardware Floating Point Unit (FPU). Further details on hardware choices are provided in [4].…”

Section: Hardware Architecturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Energy-Efficient Audio Processing at the Edge for Biologging Applications

Miquel

Latorre

Chamaillé‐Jammes

2023

JLPEA

Self Cite

View full text Add to dashboard Cite

show abstract

“…To help, we have included a guide to key SensorDrop device design considerations, potential pitfalls, and resolutions in the device failure sheet in the online repository.When combined with mature wildlife collar technologies from commercial telemetry brands, SensorDrop provides a low-risk system for researchers to innovate upon and test new sensors for wildlife research. The growing accessibility of DIY electronics components (such as SparkFun and Arduino) and the narrowing divide between wildlife researchers and technologists (facilitated by the emergence of online interdisciplinary communities such as WILDLABS) has led to a boom in researcher-developed innovations within the wildlife sciences (e.g., De La Rosa, 2019;Miquel et al, 2022;Wijers et al, 2018). SensorDrop addresses a key barrier to the continued innovation of animal-borne sensors in that it allows specific animal-borne sensors to be cost-effectively remotelydetached and retrieved from wildlife collars.…”

mentioning

confidence: 99%