Robots are becoming a standard part of manufacturing. Artificial intelligence is augmenting human intelligence to improve performance and produce everything from robotic 3D-printed smart bridges to complete buildings.
Manufacturing leads the way in the appliance of artificial intelligence (AI), changing how everything is designed, made and used.
More and more robot workers are appearing alongside their human counterparts on the production floor. This will expand as robots become capable of handling more cognitive tasks and can make real-time, data-driven decisions for themselves.
This AI revolution in manufacturing is driven by the technologies of today. From cutting unplanned downtime to predictive maintenance, better designs and simpler processes, AI-powered analytics is already helping to cut costs, improve quality and raise productivity.
Machines are quick learners
The adoption of AI in manufacturing starts with design and the new generative design technologies. These allow you to input your design goals plus any limiting factors, such as materials available, manufacturing methods and cost constraints. The computer can quickly explore all possible design solutions and test for the best solution.
The chance to think the impossible in industrial design has resulted in the first-of-a-kind 3D-printed steel bridge1 in Amsterdam. Part of its design DNA is a new IoT nervous system, with sensors throughout the structure that use Lenovo workstation-powered AI workflows to monitor the health of the bridge.
“The opportunity to examine real-time data from these sensors will provide insight to inform designs for future 3D-printed metallic structures,” says Gijs van der Velden, CEO of MX3D.
Turning downtime into uptime
One of the biggest costs for manufacturers is unplanned downtime – estimated to cost $50 billion annually2 – mainly caused by machines or systems failing.
So how do you know what’s going to fail next – and prevent it? By using machine learning and artificial neural networks (ANNs) for predictive maintenance.
Data passes through these neural networks, which learn through examples. The more training data going through the network, the more accurate the results. Teaching the network this way enables it to make the right decisions on unknown data. As the system learns, performance progressively improves.
Preventative maintenance and other automation technologies certainly pay off. In the pulp and paper industry they have improved process efficiency by 20-30%2.
Picture the quality
As consumers continue to expect the highest-quality goods, the short time-to-market and increasing regulation puts additional strains on the manufacturing process.
AI can help by spotting problems early on. For example, US-based Mark III systems are using Lenovo technologies to develop AI image recognition as part of a quality assurance program to identify product defects before shipping.
Take a walk with the robotic printers
AI is also helping manufacturing and construction to come together to enable new methods of building. Dutch architectural firm DUS (http://houseofdus.com/) and its sister company Aectual (http://www.aectual.com/company/) are using robotic printers to create the revolutionary flooring at Amsterdam Schiphol airport. Soon they expect to develop the platform using 3D software and printing to build complete homes.
This is just the start of the AI revolution that’s already bringing big benefits to manufacturing.
2 How Manufacturers Achieve Top Quartile Performance