From smart sensors to autonomous factories: why the next decade of AI will transform industry

Recently, I watched one of the many news features examining the extraordinary growth of artificial intelligence and the infrastructure required to support it. The discussion centred on the vast AI data centres being built around the world — facilities that demand enormous computing power, consume significant amounts of electricity and, in many locations, require substantial water resources for cooling.

As someone who has spent my career in industrial automation, I found myself thinking: What does this mean for the world of factory automation and industrial sensing?

More specifically, I began to wonder whether the future of AI would always require ever-larger data centres, or whether some of that intelligence could increasingly be decentralised at least in an industrial ecosystem and placed closer to where information is generated: in the sensors themselves.

Ten years ago, most industrial sensors were designed to perform one task exceptionally well. A photoelectric sensor detected the presence of an object. A pressure sensor measured air pressure. A linear position sensor reported the location of an actuator. These devices were reliable, accurate and robust, but they were fundamentally sources of data rather than sources of intelligence, with little awareness of the broader process occurring around them.

Today, that model is changing rapidly.

Modern smart sensors can provide not only measurement values but also diagnostics, health information and process insights. IO-Link enabled devices, for example, allow operators to access detailed condition monitoring information, parameter data and predictive maintenance indicators that were simply unavailable a decade ago and have ever-increasing onboard processing power.

The real excitement begins when AI enters the equation and is combined with these connected sensors.

In the not-too-distant future, many industrial sensors may include embedded AI models capable of interpreting events and making decisions locally. Rather than transmitting vast quantities of raw data to PLC controllers and external systems, sensors could process information at the source and communicate only meaningful outcomes. If successful, this shift towards edge intelligence could help reduce both network traffic and some of the computing demand currently placed on large data-centre infrastructure, while simultaneously delivering faster and more reliable automation decisions.

Having spent my career helping manufacturers solve challenges through sensing and automation, I am excited about what lies ahead. The convergence of AI, intelligent sensors and industrial connectivity has the potential to unlock levels of productivity, quality and sustainability that were once unimaginable. Achieving these outcomes will be essential if Australian manufacturing is to remain competitive, resilient and globally relevant in the years ahead. *Jim Wallace is Sales Director at Balluff AUNZ with over 30 years in the industry. He’s passionate about helping businesses improve with smart sensor technologies, especially in condition monitoring and evaluation solutions.

Top image credit: iStock.com/gong hangxu