Artificial Intelligence-Driven Robots: More Brains than Brawn

Geek and body builder playing chess

Automation and robots for manufacturing have come a long way since Unimate was introduced in the 1960’s. The machines that manufacturers are using today are smaller, safer and able to perform more than a single task without expensive programming. While these innovations have significantly increased the value that automation brings to manufacturing, what’s coming online now will transform the industry in ways that we’ve not seen since the first industrial revolution.

The 4th industrial revolution or Industry 4.0 will be built on robots that are more brains than brawn. These robots integrate physical and cognitive ability to do more than heavy, highly repetitive tasks. In the sophisticated, highly automated environments where manufacturing takes place, these behavior-based robots – fueled by new innovations in artificial intelligence (AI) – are changing the way factories are organized, operate and perform.

Building the Industrial Internet of Things from the Bottom Up

Advances in technology have always been the catalyst for transformation in manufacturing, but this time the technology is less about mechanization and physical automation and more about cognition. The Industrial Internet of Things makes it possible for manufacturers to orchestrate the production process in completely new ways. It also will automate – on a large scale – the analysis of mission-critical information in a continuous flow to enable informed, real-time decision making. It’s an exciting time, but it’s also daunting, and manufacturers are not simply going to go whole hog on rolling out an information-driven operation. They are skeptics, remember? The vision of Industry 4.0 will be achieved – in large part – by software-driven robots with innate cognitive abilities.

With AI, robots can work semi-autonomously on a much wider range of tasks. Beginning in the work cell, robots with “smarts” built-in draw from a cloud-based database of “lessons” and information to:

  • Recognize equipment and parts in a work cell and perform applicable behaviors and make “auto-complete” suggestions, e.g. recognize a tool or piece of equipment and be able to use it correctly
  • Use pattern matching to suggest error handling best practices
  • Apply a database of corrective suggestions to help task designers (that would be the robot’s human colleague) find ways to modify a task or work cell in response to a fault
  • Analyze motion profile and behavior against a global fault database to identify opportunities to optimize a task

Just as smartphones and other internet-of-things-enabled devices receive software updates that add new features and functionalities, so too will robots expand their abilities. Optimizing production at the work cell level is only the beginning. Robots will eventually share information and insight that improves performance factory-wide and ultimately, across global operations, with the ability to:

  • Learn from self and others
  • Correct self and others
  • Collect, analyze and share insights from data collected on the factory floor and from robots in other locations

Innovation: Only As Good as the Value it Provides

We’ve all seen the crazy stories about artificial intelligence and its potential to destroy life as we know it. The reality is that there is a significant value that AI-driven robots deliver to manufacturers now, and even more so in the future, such as

  • Drive continuous process improvement and improve quality
  • Reduce costs and improve margins
  • Accelerate the NPI process
  • Build factories that can produce highly customized products at mass market prices

We’re seeing customers rolling out the work cell-based model in their factories with robots at the core. How do you think next-generation AI-driven robots will change manufacturing? What opportunities and challenges do you see? Tweet me @jim_lawton.

Originally published on Forbes.