Quis-k-luz

This is the Cañón de Fernández, a riparian wetland formed by the Nazas River in the middle of the desert, and one of the most valuable ecosystems in northern Mexico. Thanks to the efforts of the nonprofit association ProDefensa del Nazas A.C., in 2003 it was designated a state reserve; in 2007, it received international recognition as a Ramsar site for its importance in wetland conservation; and in 2021, it achieved the status of a federal protected natural area, becoming a clear example of long-term commitment to conservation.

What is Quis-k-luz?

Quis-k-luz is an autonomous environmental monitoring system, developed at the Tecnológico Nacional de México, La Laguna campus, and designed to protect ecosystems at risk. It integrates acoustic, optical, and environmental sensors capable of detecting threats such as illegal logging, fires, vehicle traffic, and environmental degradation, while also recording the presence of bird species as bioindicators of ecosystem health.

The system is based on a low-power electronic platform, powered by solar energy and rechargeable batteries, enabling continuous operation in remote environments without electrical infrastructure. The collected information is processed on site and transmitted wirelessly to a monitoring network, providing real-time data for conservation decision-making.

Through this approach, Quis-k-luz aims to become an accessible, scalable, and replicable tool, supporting both local communities and institutions in the surveillance and preservation of protected natural areas.

Inspired by nature

As a curiosity, the name Quis-k-luz refers to the Quiscalus mexicanus, commonly known in Mexico as the great-tailed grackle. These birds are well known for their behavior of alerting the flock when danger is detected, with loud and distinctive calls.

In the same way, our Quis-k-luz stations act as “electronic grackles, ” detecting threats in the ecosystem and sending immediate alerts to help protect the natural environment.

How It Works?

The Quis-k-luz workflow is organized into interconnected modules that ensure real-time acquisition, transmission, and analysis of both acoustic and environmental data:

Acquisition modules

• The Acoustic Module performs audio recording and preprocessing to detect relevant sound patterns such as bird calls, illegal logging activity, or vehicular flow.
• The Environmental Module measures air quality parameters, including particulate matter, carbon dioxide, and volatile organic compounds (VOCs).

Resilient connectivity

• Both modules check for Wi-Fi availability.
• If Wi-Fi is available, data is transmitted directly to the database.
• If Wi-Fi is not available, the information is sent through the Communication Module (GPRS), ensuring redundancy in data transmission.

Processing on a Virtual Private Server (VPS)

All information is consolidated at the Database Upload stage. From there, it feeds three main analysis processes:

• Air quality analysis, including early fire detection.
• Illegal logging and vehicular flow analysis.
• Bird analysis as ecosystem bioindicators.

System outputs

• End-users access a Real-time Dashboard with raw data visualization for immediate monitoring.
• In addition, Event-based Alerts are generated only when analyses detect critical conditions, and are delivered to users through web browsers, mobile applications, or other digital channels.

In this way, Quis-k-luz integrates data acquisition, transmission, analysis, and communication into an autonomous, scalable, and reliable system for environmental conservation.

Current implementation

In the current implementation, the modules are built using M5Stack components:

• The Acoustic Module can be configured with an Atom Echo (for lower-quality audio) or with an Atom Lite combined with the Atom Echo Base with Microphone and Speaker (for higher-quality audio capture).
• The Environmental Module is based on the AirQ, designed to measure air quality parameters
• Connectivity is ensured through the SIM7080G CAT-M/NB-IoT Unit, which serves as the GPRS Module.

In the figure, the electrical schematic of the Quis-k-luz stations can be seen, consisting of the acquisition modules, the communication module, and the power management module. For power management, the Waveshare Solar Power Management Module (D) was used, a compact USB charge controller compatible with 6–24 V solar panels or Type-C power adapters. It features MPPT (Maximum Power Point Tracking) to optimize solar energy harvesting and provides a regulated 5 V/3 A output compatible with protocols such as PD, QC, FCP, PE, or SFCP. In addition, it supports lithium battery connection, includes LED status indicators, and integrates multiple protection circuits — including overcharge, over-discharge, overcurrent, and overheating — making it ideal for low-power IoT applications or autonomous environmental projects.

Backend & Data Processing

All data collected by Quis-k-luz modules is sent to a Virtual Private Server (VPS), where it is processed and stored. The backend is organized as follows:

• Bird detection: acoustic data is analyzed using birdnetlib —the Python API for BirdNET-Analyzer and BirdNET-Lite, developed by Joe Weiss— which provides a unified interface to process audio and detect species. This library leverages the Cornell Lab of Ornithology models in the backend, the same ones used by the Merlin Bird ID app, ensuring robust species recognition through deep learning. On the VPS, Python scripts execute by instantiating Recording and Analyzer objects, allowing bird identification with associated time ranges and confidence scores.
• Illegal logging and vehicular flow analysis: audio data is evaluated using pre-trained sound event models (YAMNet, PANNs, VGGish-style CNNs trained on AudioSet). These models detect patterns associated with chainsaws, engines, and wood cutting. They have been further refined by incorporating locally collected recordings and applying fine-tuning techniques on lightweight CNN/CRNN architectures, improving accuracy under real-world field conditions.
• Fire and air quality analysis: data from the environmental module feeds a rate-of-change detector that monitors sustained increases in particulate matter (PM), carbon dioxide (CO₂), and volatile organic compounds (VOCs), along with absolute thresholds indicative of nearby smoke. This analysis can optionally be combined with acoustic patterns characteristic of fire (e.g., “crackling” or “brush fire”), increasing confidence in early fire detection.
• Databases: information is stored in MySQL, managed through phpMyAdmin for accessibility and maintenance.
• Visualization: maps and interactive charts are generated using Leaflet.js and Plotly libraries, allowing georeferenced displays of air quality and biodiversity indicators.
• Processing: data handling, queries, and logic are performed with PHP scripts, which enable modular and efficient processing within the server.

This backend architecture ensures that Quis-k-luz provides accurate, accessible, and scalable data for real-time environmental monitoring.

From Ecosystems to Cities

The versatility of the proposed hardware and software architecture allows Quis-k-luz to adapt to urban environments, extending its scope beyond natural ecosystem conservation. In cities, the environmental modules are used to monitor air quality in real time, measuring parameters such as particulate matter (PM), carbon dioxide (CO₂), and volatile organic compounds (VOCs). These data are integrated into visualization platforms (see www.redknario.mx) that can be correlated with public health indicators, supporting projects that demonstrate the relationship between air quality and respiratory diseases. In this way, the system provides a flexible tool that contributes both to environmental conservation and to improving quality of life in urban areas.

Gallery

This section showcases images and videos of Quis-k-luz throughout its different stages of development and testing. It includes views of the hardware prototype, field deployments, and examples of its application in both urban and natural environments. These visual materials illustrate how the system operates in practice and support its validation for both environmental conservation and urban air quality monitoring.

Conclusions

The development of Quis-k-luz marks a significant step toward the sustainable protection of ecosystems at risk, such as the Cañón de Fernández. By integrating acoustic, optical, and environmental sensing with low-power electronics and resilient connectivity, the system demonstrates how community-based vigilance can be translated into technology, mirroring the sentinel behavior of species like the Quiscalus mexicanus.

A key strength of Quis-k-luz lies in its modularity, which makes it highly versatile. Depending on the application, the system can operate solely with its acoustic hardware for bird monitoring or logging detection, with the air-quality module for PM₂.₅, CO₂, and VOCs analysis, or with both modules combined for comprehensive monitoring.

Powered by solar energy and designed with redundant data transmission, the platform ensures continuous operation in remote areas without infrastructure, while also adapting to urban contexts. In cities, Quis-k-luz links environmental data with public health indicators, supporting studies on the relationship between air quality and respiratory diseases.

Altogether, Quis-k-luz stands out as an accessible, scalable, and replicable tool that strengthens conservation strategies, supports scientific research, and encourages community participation in environmental stewardship. Its innovative and adaptable design positions it as a reference model for future monitoring projects in Mexico and beyond.