SoundBox QT-2100AI
AI Smart Sound Box
The QT-2100AI is a smart sound monitoring device developed by ZOTECH. This product provides real-time, high-precision sound pressure monitoring, AI event classification or identification, long-term recording, and diverse communication integration features. Integrated with the MagicBrain AI accelerator, it can monitor sounds and events, including vehicle noise, construction site noise, bird calls, and various environmental sounds.
The product is meticulously designed for smart cities, smart construction site management, and ESG ecological monitoring applications. It can perform sound feature extraction and classification inference at the edge, effectively reducing cloud computing load and significantly improving system response speed. This edge computing architecture not only lowers network bandwidth requirements but also ensures data privacy and system reliability, continuing to operate even in environments with unstable networks.
Specifications
Modular System Architecture: Leading High-Efficiency and Flexible Acoustic Monitoring
The QT-2100AI adopts an advanced modular system architecture, consisting of five core components: the audio sensing unit, main control MCU, AI acceleration module, diverse communication modules, and power management system.
Audio Sensing Unit
Integrates MEMS and analog dual microphones with low-noise design, ensuring high-precision audio acquisition across a wide frequency range, effectively capturing faint or sudden sounds.
Power Management
Supports wide voltage input (DC 8V-36V) to ensure stable operation. It also supports efficient charging and solar power integration to extend battery life.
Diverse Communication Modules
Built-in Wi-Fi / BLE, with optional communication interfaces such as 4G LTE, LoRa, NB-IoT, or RS485, ensuring stable, efficient, and secure remote data transmission according to different field requirements.
Main Control MCU
Utilizes a low-power processor, supporting timed wake-up and multi-task scheduling, effectively reducing power consumption and memory requirements, extending device deployment time, and performing real-time audio pre-processing and data compression to optimize transmission efficiency.
AI Acceleration Module
Equipped with a MagicBrain AI accelerator, based on a high-performance Cortex-M55 processor and Ethos-U55 NPU architecture, it fully supports TensorFlow Lite model execution. Empowered by edge computing, the QT-2100AI can perform real-time acoustic analysis directly at the data acquisition site, significantly reducing data transmission latency, saving cloud processing costs and bandwidth, and enhancing data privacy.
Six Core Feature Highlights
1
Real-time AI Sound Event Recognition
Offers a variety of sound classification or recognition models to choose from, including car/motorcycle noise, construction sounds, bird calls, frog croaks, environmental background sounds, etc. Products with the most suitable corresponding model can be selected according to different application scenarios to achieve the best intelligent effect.
2
Supports Dual Microphone Stereo Recording
Utilizes dual MEMS + analog microphone input paired with advanced noise reduction algorithms, effectively filtering wind noise and environmental interference, greatly improving event recognition accuracy. Even in complex acoustic environments, it can precisely identify target sound events.
3
Smart Dual Trigger Mechanism
Supports a dual trigger logic of dB threshold and AI event classification. It can be set to automatically start recording and data upload when the sound pressure exceeds a specific threshold and/or a specific sound event is identified, ensuring that important events are not missed.
4
OTA Remote Update
Built-in Over-The-Air firmware and AI model update mechanism, supporting remote operation and model optimization. System upgrades can be completed without on-site personnel, significantly reducing maintenance costs and enhancing system flexibility.
5
Diverse Communication Integration
Built-in Wi-Fi and BLE, with optional LoRa, NB-IoT, 4G, and other modules. The most suitable communication solution can be selected according to different field network environments, ensuring the stability and reliability of data transmission.
6
Platform Integration Capability
Seamlessly integrates with GIS / IoT platforms for real-time sound map generation, hotspot analysis, and long-term statistical report generation, transforming raw sound data into valuable visual information for decision-making.
Versatile Communication Options, Easy Integration with Various Systems
SoundBox's diverse communication solutions are key to its flexible system integration. By supporting multiple communication protocols, SoundBox can adapt to the network demands of various deployment environments. Whether in high-density urban areas or remote monitoring points, the most suitable connection method can always be found.
WiFi / BLE
Short-range, high-frequency communication options suitable for applications with high data volume and frequent data transmission, such as indoor building monitoring points or urban hotspots.
RS485
A common communication protocol in industrial automation, providing stable and reliable wired connections, especially suitable for integration in factories or large facilities.
LoRa
Long-range wireless communication technology, low power consumption, and wide coverage, making it an ideal choice for remote areas or applications requiring long-distance transmission.
4G/NB-IoT
Extended support for narrow-band Internet of Things (IoT) technology, utilizing existing telecommunication network infrastructure to provide wide coverage and stable connections.
This communication flexibility enables SoundBox to be easily integrated into existing infrastructures such as smart streetlights, building management systems, or security platforms. System integrators can select the most appropriate communication mode based on project requirements, balancing data transmission speed, power consumption, and cost to achieve an optimized system architecture design.
Industrial-Grade Design, Meeting Diverse Environmental Challenges
SoundBox's industrial-grade design is reflected in every detail. From material selection to structural design, it undergoes rigorous testing and validation. Its durable casing not only provides excellent physical protection but also considers critical factors like heat dissipation and moisture resistance, ensuring stable operation of internal electronic components in extreme environments.
IP67 Water and Dust Resistance
Optional IP67 protection ensures the device operates normally in harsh outdoor environments, resisting rain and dust intrusion.
Solar Power Supply
Optional solar modules provide self-sufficient energy, suitable for deployment in remote areas with inadequate power infrastructure.
Low Power Consumption Design
Supports multiple low-power operation modes, extending battery life through intelligent wake-up mechanisms, reducing maintenance frequency and costs.
Strong Temperature Adaptability
Special design allows the device to operate stably in extremely high or low-temperature environments, adapting to Taiwan's variable climate conditions.
These characteristics make SoundBox particularly suitable for deployment in various challenging environments, such as outdoor construction sites, remote mountainous areas, coastal regions, or unattended monitoring points. Whether exposed to sunlight for extended periods or facing wind and rain, the SoundBox QT2100 series can continuously provide stable sound monitoring services.
One-stop AI Perception System Integration Service
SoundBox provides a complete one-stop AI perception system integration service, forming a closed-loop solution from hardware equipment to software platforms and system integration. The modular hardware design allows customers to flexibly configure according to specific needs, while standardized SDKs and APIs greatly simplify the development process.
Modular Hardware
Complete solution for Sensor + AI + Communication Interface
SDK / API
Standard JSON format data transmission interface
Rapid PoC
Quick verification with development boards and pre-trained models
Cross-Platform Integration
Seamless integration with SCADA, IoT middleware, and cloud platforms
GIS Visualization
Full support for Geographic Information Systems
For customers who need rapid concept verification, SoundBox provides development boards and pre-trained model kits, allowing teams to conduct tests in the shortest possible time. In terms of system integration, SoundBox supports seamless integration with various SCADA, IoT middleware, or cloud platforms, and has specially optimized its connection capabilities with GIS systems to achieve precise geographical information visualization.
"Finding Urban Sanctuaries" Project — Hackathon Grand Prize
"Finding Urban Sanctuaries" project is a successful case of SoundBox in practical application. Led by a team from Academia Sinica, combined with SoundBox technology, it created a comprehensive "Smart Urban Noise Map" platform. This platform collects noise data from various urban areas through a sound sensor network and performs analysis and visualization.
This innovative project stood out in the 2024 Presidential Hackathon, winning the grand prize. It fully demonstrated SoundBox's immense application potential and social value in policy analysis and urban environment improvement. The project results not only provide urban planners with important decision-making basis but also offer residents an intuitive reference tool to understand the sound quality of their living environment.
Hackathon Honor
Top five in the 2024 Presidential Hackathon, showcasing SoundBox's value in policy analysis and urban environmental improvement.
Urban Noise Problem
Urbanization leads to increasingly severe noise pollution, affecting residents' quality of life and health, requiring innovative solutions.
SoundBox Solution
Deploy a sound sensor network to collect noise data from various urban areas, establishing a comprehensive sound database.
Smart Noise Map Platform
Visualizes sound data analysis results, helping identify urban areas that require noise mitigation measures.
Diverse Application Scenarios
Smart City
Through environmental sound AI classification, accurately map the city's soundscape and conduct long-term acoustic data analysis. This will effectively support urban planning and noise control policy formulation, significantly improving citizens' quality of life.
Smart Construction Site Management
Combining noise decibel thresholds with AI construction site noise recognition, the system automatically triggers monitoring, instantly detects abnormal construction sounds and issues alerts, and provides recording for evidence, effectively assisting in safety management and dispute resolution, ensuring construction compliance.
Healthcare
For wards and care centers, accurately identify abnormal sounds such as calls for help and screams, promptly warn of emergencies, and effectively ensure the safety of patients and the elderly.
Safety Protection
AI sound recognition can identify common sounds like glass breaking, emergency alarms, and abnormal structural impacts, as well as a wider range of sudden incidents such as gunshots and physical altercations. Through 24/7 acoustic monitoring, the system effectively compensates for the detection limitations of traditional visual monitoring in blind spots, low-light environments, or obscured areas. Once a potential threat is identified, it immediately triggers alerts and automatically records evidence, significantly shortening response time and comprehensively enhancing the safety protection level and intelligent response capabilities of the area.
Ecological Environment Monitoring
Through AI bird recognition models, accurately identify different bird species, continuously record and store them in a biodiversity soundscape database, and integrate environmental data such as time and geographical location to provide quantitative biodiversity indicators, helping to assess environmental health and ecosystem service functions, and supporting ESG sustainable development.
Case Study - Searching for Urban Pure Land
Origin Technology partnered with Academia Sinica and government agencies to create urban 'Quietness Suitability Index (QSI)' maps using QT-2100AI Sound Boxes and big data analysis.
Sound sensing nodes were strategically deployed across New Taipei City, establishing noise spectrograms and decibel databases. Through continuous data collection and machine learning, the QSI map accuracy was refined, enabling citizens to find peaceful spaces while providing scientific urban planning insights.
This low-cost, large-scale deployment monitors real-time sound environments, reveals long-term noise exposure health risks, and delivers practical quiet zone information to improve urban quality of life.
Case Study - Ecologyse
Taipei Guandu Academia Sinica Ecological Pond Monitoring
Biodiversity conservation requires long-term biological population monitoring to track ecological changes and guide conservation measures. Sound serves as direct species evidence and a real-time biodiversity indicator, overcoming traditional manual observation limitations in time, scope, and efficiency.
This valuable sound data supports scientific research, helps government agencies and enterprises achieve ESG sustainable development goals, and aligns Taiwan's ecological conservation with international standards, preserving natural assets for future generations.
Complete Accessories & Expansion Solutions
The QT-2100 boasts exceptional environmental adaptability, capable of fully meeting the stringent demands of various deployment scenarios through a diverse range of dedicated accessories.
Protective Installation
IP67 industrial waterproof casing and various mounting brackets (wall/lamp post/ceiling)
Power Module
Supports lithium and lead-acid batteries, providing stable and reliable power to ensure the device operates independently for extended periods in off-grid environments without mains power.
Solar Power Supply
The solar charging kit ensures the device can operate autonomously and sustainably for extended periods in any location without power, maximizing monitoring efficiency and data continuity.
Communication Expansion
Supports RS-485 wired expansion, facilitating connection to multiple sensors or devices. Additionally, with LoRa or NB-IoT wireless communication converters, wide-area deployment can also be achieved.
Environmental Sensing
Supports additional sensors for PM2.5, gases, temperature, and humidity, expanding the monitoring range and providing more comprehensive environmental data collection capabilities.
Whether you are an urban planning contractor, system integrator, or ecological research institution, SoundBox offers an innovative sound sensing solution to help you create smarter, more livable urban environments. Please contact us via the following methods:
Website：https://www.zot.com.tw/
Email：[email protected]
Address：2F, No. 5, Alley 22, Lane 513, Ruiguang Road, Neihu District, Taipei City