Robotics & A.I. - Artificially Intelligent Cobots in Automation

Originally published on July 2nd, 2018.

Who you are defines how you think of robotics and automation.  Software experts and IT may think of internet bots.  They might also think about the new, emerging field of Robotics Process Automation (RPA), which is software that can do mundane and administrative computer tasks.  RPA reduces repetitive tasks such as checking, verifying and transferring data.  Manufacturing facilities will think about physical robots or cobots that are also deployed to handle repetitive tasks such as loading and unloading a CNC machine or installing a computer cover.  They can also be used to automate dangerous tasks such as lifting, welding or removing paint.

What is common between the software robot and the physical robot?  The answer is advancements in Artificial Intelligence (AI).  AI is the field of creating intelligent machines that work and react like humans. For example a large UK utility case study illustrated how 2 people managing 300 RPA robots performed 300 million transactions per quarter.  Imagine no longer having to do mundane transcribing, checking or sorting through email!  Many people ask – does this mean the workforce will decrease?  The answer has been clear – no.  With the data explosion (call it “big data”) the influx of information bombarding the worker is unprecedented – I regularly see inboxes with 1,000 to 10,000 emails in it.  These workers – myself included - can no longer keep up (this has led to the phrase “email is dead” because few can effectively manage their inbox).  AI can then weed through the noise and help the worker focus on the most important tasks.

These advancements in AI contribute to advancements in a physical environment.  As cobots get smarter on how to interpret the world around them and decide what action to take next they become easier to implement.   Users don’t need to know software or programming languages – they simply show the robot what to do.  The next step in this evolution is machine learning so the robot can see the environment, interpret the inputs, and make decisions on the next step to take.  FANUC and KUKA are each working on implementing AI and machine learning to reduce or eliminate the traditional extensive programming.  The implementation of this in the real world means explosive and broader use of robots.  For example, cobots can be easily deployed to load and unload a CNC today.  However, this still requires material to come to the robot in a predictable way.  Machine learning could expand the robot’s functionality – it could recognize that it is out of parts to load into the CNC and retrieve them.  Robots can interact with other robots and AGVs to accomplish tasks.

The same question applies here – does this mean the workforce will decrease?  The answer is still no.  Job needs and required skill sets will shift.  The worker will be relieved of the increasing quantity of necessary but mundane tasks, checks and validations – allowing a portion of their day to go to creativity, implementing improvement ideas, preventative maintenance or simply being able to focus on doing their job and doing it well.   Who you are defines how you think of robotics and automation.  Whether it is Robotics Process Automation or Machine Learning it has the potential to have a very positive effect on your daily work life.