Artificial Intelligence (AI) is by no means a new phenomenon. For decades we have been talking about AI as a technology with the potential to radically disrupt our society and impact the future of mankind – some being optimistic, others dystopic in their view on AI.
But with the emergence of technologies like ChatGPT, it increasingly looks like the future is now and it’s hard to think of a time when AI has been a hotter topic than the present.
One of the key reasons for AI now being discussed in workplaces all over the world is because computer processing has taken giant leaps forward in recent years. We now have the processing power to handle the vast amounts of information and data required for AI technology, something we simply didn’t have before.
This hardware development has paved the way for an AI breakthrough, including new software programs like ChatGPT. But while chatbots are still in the early stages of transforming how we communicate and gather information on the internet, AI technologies are already making a real-life difference in another arena – manufacturing.
Here are four examples of how AI is impacting industrial automation today and how it can benefit manufacturers all over the world by making it simpler than ever to automate complex and diverse tasks, even in unstructured environments.
Humans can look at disordered objects –such as parts in a bin - and immediately see the difference and understand which of them can be handled without interfering with other objects. Automation engineers know that this isn’t always the case for robots, to put it mildly. As a result, bin-picking of unstructured items has traditionally been thought of as a notoriously difficult problem to solve. But this is changing with AI.
Take for example Apera AI’s ‘4D Vision’ technology, which is challenging the status quo by providing collaborative robots (cobots) with “humanlike perception” – a claim that sounds hyperbolic at first but is borne out on several levels and enables faster, more effective robot performance, especially on bin-picking applications.
With the use of scanners and cameras, ‘4D Vision’ can identify the “most pickable” objects and inform the cobot of the fastest and safest path to handle them. The cobot is provided with pose estimation and path planning data ensuring that the robot takes a safe, collision-free path to accomplishing its goal.
The mainstream understanding of AI is a technology that’s able to “think” by itself and make decisions without prior teaching or instructions. Even though this is not always the case, these upsides are exactly what you get with the robobrain.vision kit from Robominds designed for the logistics industry for e.g. kitting, order picking or de-palletizing tasks. To see it in action, see this video.
Simplified, most automation solutions within manufacturing are programmed to handle a specific object with set dimensions. Of course, these solutions can be programmed to handle further variations, but they rely on humans telling the robot what objects to handle and what to do with them.
With this kind of camera-based AI technology, the robot can pick up different objects regardless of their shape or size. And by not having to spend time on teaching or programming the robot, customers are given even greater flexibility and can change the objects being handled without spending time on re-programming.
Another example of how AI enables industrial robots to deal with variance in position, shape or movement is MIRAI from Micropsi Industries.
Instead of being dependent on specific measurements being pre-programmed, it’s able to generate robot movements in real-time. This means the robot can do e.g. assembling, gripping, screwdriving or testing tasks, even if the position of machines or objects fluctuate from time to time.
Inbolt's AI-based Inbrain is another technology using AI to handle variations and moving parts. It processes massive amounts of 3D data at high frequency and identifies the position and orientation of a workpiece, adapting the robot trajectory in real-time, which makes it ideal for assembling, handling, finishing and testing. To see it in action, see this video.
AI can also be used to give robots the touch senses of robots. The AI control software from AICA enables the robot to learn precise tasks such as assembly of gears, even when the task varies every time.
Another important upside of AI in industrial automation is that its constantly improving - automatically.
The more your robot is working, the more data the AI application is gathering, and with this data the underlying algorithm can continously optimize, adjust and improve the robot’s performance.
This level of self-learning means that, as a customer, your automation solution will improve by the day – without you having to spend time and money on updates or upgrades to your solution.
The upsides of this new synergy between AI and cobots are evident. Manufacturers looking to cobot automation to overcome their business challenges – whether its labor shortage, improving employee’s wellbeing or raising quality or productivity – are now able to solve extremely complex tasks, even in unstructured environments. And, at the same time, AI products are presenting manufacturers with an unforeseen level of flexibility and simplicity, as well as raising quality and reliability.
It's clear AI is already making an impact on industrial automation, but the best part is we’re only just scratching the surface. The future might be here, but the best it yet to come.
Anders Billesø Beck is the Vice President Strategy & Innovation, Universal Robots.