Enterprise AI Analysis
Artificial Intelligence of Things: A Survey
The proliferation of the Internet of Things (IoT), such as smartphones, wearables, drones, and smart speakers, as well as the gigantic amount of data they capture, have revolutionized the way we work, live, and interact with the world. Equipped with sensing, computing, networking, and communication capabilities, these devices are able to collect, analyze, and transmit a wide range of data including images, videos, audio, texts, wireless signals, physiological signals from individuals, and the physical world. In recent years, advancements in Artificial Intelligence (AI), particularly in deep learning (DL)/deep neural networks (DNNs), foundation models, and Generative AI, have propelled the integration of AI with IoT, making the concept of Artificial Intelligence of Things (AIoT) a reality. The synergy between IoT and modern AI enhances decision making, improves human-machine interactions, and facilitates more efficient operations, making AIoT one of the most exciting and promising areas that have the potential to fundamentally transform how people perceive and interact with the world.
Executive Impact & Breakthroughs
The overarching goal of this survey is to provide a systematic and comprehensive review of AIoT research. As shown in Figure 2, we organize the literature of AIoT in a taxonomy consisting of four main categories: sensing, computing, networking & communication, and domain-specific AIoT systems. Specifically, Sensing: Sensing serves as the foundation of AIoT. In Section 2, we survey AI-empowered sensing mechanisms and techniques in AIoT that cover research directions related to motion sensing, wireless sensing, vision sensing, acoustic sensing, multi-modal sensing, earable sensing, and Generative AI for sensing (Table 1). Computing: Computing is the brain of AIoT. In Section 3, we survey fundamental compute tasks that lie at the core of AloT, covering topics related to on-device inference, offloading, on-device training, federated learning (FL), and AI agents for AIoT. Networking & communication: Networking & communication serve as the backbone of AIoT. In Section 4, we survey AI-empowered networking & communication techniques related to a variety of networks including cellular/mobile networks, Wi-Fi networks, visible light communication (VLC), and LoRa (long range)/LoRaWAN (long-range wide area network). – Domain-specific AloT systems: The advancements in sensing, computing, and networking & communication lay the foundation for the development of AIoT systems designed for specific application domains. In Section 5, we survey these AIoT systems in important application domains including healthcare and well-being, video streaming and analytics, and autonomous driving, as well as AR, VR, and MR. We have established a GitHub repository to organize the papers featured in the survey at https://github.com/AIoT-MLSys-Lab/AIoT-Survey. This repository will be actively maintained Although there are several surveys on topics relevant to AIoT [21, 30, 89, 94, 168, 196, 233, 324, 329], they focus on some specific aspects of AIoT. In contrast, this survey provides a holistic view of AIoT research. More importantly, we primarily focus on literature on sensing, computing, network-ing & communication, and domain-specific AloT systems that are built upon modern Al techniques such as DL, foundation models, and Generative AI. It is our hope that this survey along with the GitHub repository could serve as valuable resources to help researchers and practitioners gain a comprehensive understanding of AIoT research and inspire them to contribute to this important and exciting field.
0
Activity Classification Precision
0
Activity Indexing Recall
0
3D Human Pose Error
0
MAC Frame Reception Rate Increase
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Motion Sensing
Motion sensing involves using IMU sensors to capture various motions like arm postures, body movements, and physical activities. AI-empowered techniques enhance human activity recognition (HAR) and arm tracking.
0
Precision in Activity Classification (Lasagna)
AI-Empowered Motion Sensing Pipeline
Sensors: Accelerometer, Gyroscope, Magnetometer
→
Sensed Data: Acceleration, Angular Velocity, Magnetic Field
→
AI Models: CNN, RNN, BILSTM, GRU
→
Sensing Tasks: Human Activity Recognition, Arm Tracking
Unsupervised HAR (Lasagna)
Liu et al. [160] introduce Lasagna, an unsupervised learning based HAR framework that extracts common bases of human motions in an unsupervised manner, creating a universal multi-resolution representation for common human activities. This addresses the limitation of existing HAR frameworks needing pre-defined activities and prior knowledge or labeled data for supervised training.
Results: The prototype system achieves 98.9% precision in activity classification and nearly 100% recall with about 90% precision in activity indexing.
Wireless Sensing
Wireless sensing uses RF signals (RFID, Wi-Fi, mmWave, LTE, LoRa) for contact-free object and individual sensing, grouped by frequency bands. It covers tasks like HAR, 3D human mesh construction, indoor localization, and object recognition.
| Technology | Advantages | Limitations |
|---|---|---|
| RFID Sensing |
|
|
| Wi-Fi Sensing |
|
|
| mmWave Sensing |
|
|
AI-Empowered Wireless Sensing Pipeline
Sensors: mmWave Radar, RFID Tag, Wi-Fi Router
→
Sensed Data: RSSI, Phase, SNR, AoA, ToF, CSI, Power, Amplitude, Frequency, Doppler Frequency Shift
→
AI Models: CNN, VAE, LSTM, cGAN
→
Sensing Tasks: Localization, Human Activity Recognition, Imaging, 3D Human Mesh Construction, Gesture Recognition, Voice Reconstruction
0
MAC Frame Reception Rate Increase (AiFi)
Calculate Your Potential ROI
Estimate the efficiency gains and cost savings for your organization by integrating AIoT solutions, tailored to your industry and operational specifics.
Annual Savings
$0
Hours Reclaimed Annually
0
Your AIoT Implementation Roadmap
Our phased approach ensures a smooth transition and maximum value realization from your AIoT investment.
Phase 1: Discovery & Strategy
Conduct a comprehensive audit of existing IoT infrastructure and data. Define key AIoT use cases and align them with business objectives. Develop a customized AIoT strategy and technology roadmap.
Phase 2: Pilot & Proof of Concept
Implement AIoT solutions in a controlled environment. Validate technical feasibility and demonstrate initial ROI. Gather user feedback for refinement and optimization.
Phase 3: Scaled Deployment
Expand successful pilot projects across the enterprise. Integrate AIoT with existing IT systems and workflows. Establish governance, security protocols, and performance monitoring.
Phase 4: Optimization & Future-Proofing
Continuously monitor and optimize AIoT system performance. Explore advanced AI/ML models and new sensing modalities. Stay ahead of industry trends and evolving technological advancements.
Ready to Transform Your Enterprise with AIoT?
Our experts are ready to help you navigate the complexities of AIoT and unlock its full potential for your business. Book a complimentary strategy session today.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat