There are many components to cognition, what humans consider “thinking”. There are parts scattered around research areas and markets. Too many times, people are focused on one specific branch. What’s more interesting to see is how the parts are interacting. Manufacturing is a clear area where the integration of different parts of artificial intelligence (AI) is happening.
People are always asking “What is artificial Intelligence?” The simple answer is “Whatever computer folks still don’t know about cognition.” As soon as an area is well enough understood, that becomes a discipline and AI remains whatever is left. That lead many pundits to continually claim that AI is a failure because it hasn’t yet solved everything. In reality, as hardware and software become powerful enough to support each area, it was able to grow; but the AI heritage remains.
Today, there are four key areas of AI achieving advances and market penetration:
- Robotics: engineering and science blended to make physical constructs that move and act
- Vision: systems that recognize objects based in visual input
- Automated Neural Networks: combinations of nodes of computing that, combined, that provide a method of analysis very loosely based on human thought processes
- Natural language: the capacity to both understand and create communications, both written and spoken
All the areas are not only used in manufacturing but are overlapping to provide coherent and impressive systems.
AI On The Manufacturing Line
Robotics is the part of AI that has been in manufacturing the longest. As far back as the 1950s, engineers worked with manufacturing groups to create robotic arms to enhance production. Today, robotics is a successful industry with a strong presence. According to Loup Ventures and the International Federation of Robotics, the robotic market value was $11.1B in 2015 and is expected to grow to $33.0B by 2025.
That is, in part, due to inclusion of vision and other aspects of AI. The ability to improve vision systems has been used primarily in quality control. Does a widget look like it is within tolerance limits, or does it have a defect? More complex, faster processing, vision advances have meant that the inspection can move upstream in the manufacturing process. Rather than waiting for a final inspection on a conveyor with a fixed camera, robots can continually check parts and flag defects during the product cycle, improving ROI by rejecting defective products in the process rather than working on other aspects of each item for wasted energy and materials.
Improvements in vision have also helped in the development of collaborative robotics. The history of robotics in manufacturing is to have an individual arm do work, pass the item down the production line, have another arm do work, and repeat as needed. Vision, and other technologies, are helping to create assembly lines where more than one machine can work on an item at a time, working in a more flexible manner to avoid collisions. That means a faster manufacturing cycle.
Automated neural networks (ANN) are advancing robotics both as a source for improvements in computer vision and with collaborative computing. The ability of ANNs to learn complex tasks and interactions has meant the rapid advancement in vision skills seen in the last decade. In addition, collaborative computer requires the adaptive ability that that ANNs can provide.
Recent advances in natural language processing (NLP), as seen with Amazon Alexa and
AI In Manufacturing – Not Only On The Line
The above discussion of the four key types of AI used examples focused on the manufacturing line. The potential, however, goes far past that.
On the manufacturing floor, robotics can help supply raw material from warehouse to each position on the line. Those robots will use all the aspects of AI to navigate from warehouses to the appropriate line positions.
Logistics, supply chain, and other processes that support manufacturing will also see advances due to the use of AI. That will positively impact bottom line costs and responsiveness to customer demand.
Over the last decades, robotics has become a normal site in manufacturing. Advances in AI mean that the next two decades will even larger improvements in manufacturing and a continued increase in market penetration.
