gshc2020.com

Today, AI largely operates on existing digital data. Its next evolution will be driven by the creation of new digital data generated from the physical world, captured through sensors, devices and systems deployed at a massive scale. That shift is where IoT, and particularly LPWAN, becomes foundational rather than complementary.

AI Needs A Nervous System

Thinking about AI as a “digital brain” primarily trained on pre-existing data makes it easier to understand. However, to perceive and influence the physical world, AI needs more than a brain. It needs sensor inputs and actuator outputs.

This is where IoT is required. It functions as the nervous system, connecting the brain to the body, collecting signals from the environment, interpreting them and enabling decisions that translate into physical action through devices and machines.

The challenge is scale and diversity. There are thousands of sensor types: vibration, acoustic, thermal, environmental, vision and location, among others. They all face similar constraints that must be addressed in product design, including long-range, ultra-low power, low cost, intermittent connectivity and security. Designers must decide where AI processing occurs, whether in the cloud, at the edge (gateway) or on-device (sensor-level ML). The disparate set of requirements and functional demands calls for a highly versatile central nervous system that is flexible enough to support the Massive IoT space.

Looking at Massive IoT requirements, billions of battery-powered devices must operate across vast areas with minimal maintenance. LoRaWAN LPWAN technology addresses these requirements by combining long range, ultra-low power consumption, low total cost of operation and flexibility across public/private/satellite networks. It handles the heavy lifting for massive, sparse, low-power deployments while working alongside other wireless technologies such as Wi-Fi, BLE and cellular, each contributing specific capabilities suited to specific applications.

The AI To IoT Flywheel

The AI-to-IoT flywheel accelerates innovation. First, IoT feeds AI with richer, real-time and more diverse data, improving model accuracy and relevance. Then, AI feeds IoT with better models that reduce data volume, improve battery life and enable predictive maintenance and autonomous control.

The result is a reinforcing cycle. More sensors produce better models. Better models enable more intelligent filtering and lower operating costs. Lower costs allow for broader deployments, which in turn generate more data to develop even better models. The flywheel continues to turn, driving continuous improvement across both domains. As this cycle matures, AI enables further acceleration across robotics, advanced data analytics, satellite IoT backhaul, digital twins and sustainability.

AIoT Will Drive Massive IoT

In AIoT technology, as the name suggests, AI and IoT work together to collect data at the edge, closer to the source, and process and act on it there. This shift toward edge intelligence is gaining momentum across industries.

As IMARC Group explains, “The magnifying expansion of IoT devices in a wide range of sectors is a major trend boosting the U.S. edge computing market. Various businesses are actively opting for edge computing to process data closer to IoT endpoints, improving real-time decision-making and minimizing latency. Furthermore, this is particularly crucial for applications in healthcare services, smart cities, and industrial automation, where rapid data processing is requisite.”

Massive IoT is where AIoT will become mainstream. The market is enormous, and innovation is everywhere. And ultimately, as the main nervous system, LPWAN wireless networking technologies hold the key to unlocking the true potential of AIoT at scale.

Several concrete innovations are already emerging. Event-driven edge ML sensors, such as vibration sensors from manufacturers like Honeywell and Advantech can detect bearing faults, imbalance or misalignment. These devices process raw sensor data locally and, instead of streaming large volumes of data to the cloud, they run ML and only deliver events. Vision/image sensors from companies like Seeed Studio and others are another emerging technology, as they perform on-device image processing for functions like object detection, occupancy and safety compliance, then send metadata/events instead of frames.

Additionally, Kudzu Technologies has brought to market another application that uses AIoT to analyze network behavior and detect anomalies such as misconfigurations, interference or suspicious activity, improving overall network reliability and security.

Bringing AIoT To The Physical World

So, what are the real-world applications for this technology? With analysts predicting massive shifts towards edge and on-device AI processing, where will it manifest in the physical world? Some of the leading applications include:

• Using ​predictive maintenance via vibration events and automated work orders to enhance industrial automation.

• Occupancy-driven HVAC/lighting control, leak detection and elevator monitoring could make buildings smarter and more efficient.

• Water metering, gas monitoring, streetlighting and waste management will improve our cities and utilities.

• Soil moisture events, livestock monitoring and wildfire sensing will make farms more productive and strengthen our environmental resilience.

• Geofenced events ​for pallets/containers and tracking cold-chain exceptions support logistics application and strengthen supply chains.

• PPE detection events and air quality thresholds triggering ventilation or gas shut-off support safety and strengthen compliance with government regulations.​

Bringing Massive IoT To Life

Although AI’s uses beyond generative engines have not yet reached the mainstream, positive IoT applications are already demonstrating their value, driving the current ramp in Massive IoT. With the industry's current pace of innovation, leveraging AI and IoT is bringing within reach the ability to solve critical challenges impacting people's quality of life and sustainability.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?