AI in Manufacturing Industry Comes With AI in Designs

Manufacturing industry is in a transition stage taking big leaps towards digitization and automation; resulting in what we today call as Industry 4.0. Some of the aggressive users of robotic technology and artificial intelligence – something most players are still grappling to adopt in the manufacturing industry – see artificial intelligence the next big logical step for improving the productivity.

Originating from Information Technology and Software industry, AI has got benefits in stores for numerous others and manufacturing is just one among them. But to explore the true potential of AI in manufacturing there has to be the inculcation of AI for manufacturing right from the concept ideation of product design which helps in better streamlining of processes at a broader level.

According to a report by Infosys, about 29% manufacturers have adopted AI as a part of their operations and decision-making process. Operations and decision making processes necessarily mean that the manufacturers leverage sensor technology, data analytics, and 3D printing, and digital twinning as well, to a certain extent.

The report cites a wonderful example of a car maker who is running low on a part or a component count having software that can generate the order and issue new parts. In case if the part an entirely new design, design teams will receive generate and dispatch orders; and the design teams can quickly and neatly develop these parts using sophisticated technologies of 3D printing originating as a result of 3D CAD modeling. Such quick response can reduce manufacturers work to hours which originally took days.

This is an explicit example of how AI has transformed not just shop floor core operations but everything is automated and taken care of very efficiently, very quickly.

AI is a big transformation

With acceptance of deep learning in AI, each stage of the manufacturing process is being monitored by sensors and data fetching devices generating humongous data. This data is then used along with machine learning algorithms and software in AI to run the day to day operations like part ordering, BOM release, initiating a purchase or design orders, fetching the part from inventory or orders to dispatch the part from shop to the warehouse.

It drives valuable insights to streamline and uplift the quality of final merchandise being manufactured in a particular unit. From inflexible production setups and assembly lines to managing capacity along with rising production costs and from variable quality yields and inability in managing shop floor operations, all the issues are easily addressed with AI.

Digital Twin; best application of AI for manufacturing

As the manufacturing industry propels towards big data and unfolds methods of gathering more data with each passing day, digital twin concept is going to take the center stage as an application of AI. A virtual model of an actual physical object can essentially be used for computing intelligent and critical processes along with information backed decision-making processes which can hardly result in a blunder.

By continuous monitoring of such digital twin models, design engineers can essentially shorten the design cycle times by gaining design insights beforehand the modeling breaks the ground. Entire product lifecycle can be enhanced, streamlined the maintenance and hone manufacturing shop floor processes by appropriate arrangement of machine shops in a logical order.

On the other hand, for core manufacturing operations, where physical machines are installed, behaviors can be compared and maintenance and downtimes can be predicted with the sensor-based data.

Bringing AI to manufacturing is bringing AI to designs

Essentially a digital twin is created and when physical parts are connected to sensors sending signals to these digital prototypes that have been synthesized. The data so gathered from manufacturing and operations sites are fed to data analytics software and AI for gaining insights to optimize designs for achieving maximum operational efficiency.

This process repeated over several times business operations at design level, manufacturing, operations, and management level can be uplifted sharply and the digital twin model can act as a live model.

Likewise accurate processes and product design requirements being generated from the shop floor, even CAD design engineers are empowered to address the manufacturing needs with due diligence. Furthermore, when final CAD models of the products are built, they become the assets for turning the data into digital prototypes of the existing physical part.

Rationale for AI in designs

We say, “AI should occupy a significant place in designing”, is based on the fact, however weak, is the higher acceptance of generative designs in several recent years. The next wave of automation is sure to impact manufacturing industry for every stage from designs to manufacturing. As a consequence of which, product designs will be heavily disrupted as we see several generative design tools by Autodesk and Dassault Systemes – the two leading enterprises and most prominent of CAD software developers.

Another instance is the fact that feature and character recognition which are a part of AI for years are also a part of SolidWorks. In fact, these features are so deeply embedded into the CAD platform that they aren’t even recognized as AI anymore and thought of as a CAD feature only. On the other hand, generative designs in AutoCAD and Dreamcatcher project by Autodesk are noteworthy accomplishments in proving the fact that AI has to be inculcated in designing for efficient leverage in manufacturing.

Time to bring AI onboard for efficient delivery

As we, the professionals of design and manufacturing industry, are exploring automation and digitization to unlock the full potential of our capacities, we should embrace AI with open arms for the product lifecycle right from concept initiation to manufacturing processes and everything in between, essentially the designs.

For every product design, industrial machinery to commercial products, all designs are getting generative and digital twin concept of AI is adding value for manufacturing and operations of these products. Designers are looking for shorter design development cycle and manufacturers are looking for efficient shop floor processes – both these can come from AI applications on whole.

The report cited above shows that about 1% of manufacturers surveyed use AI technologies for more than six months while 22% use it for a time period that is between six months and a year. Surprisingly only 2 % of manufacturers use it for 5+ years and highest number about 58% use it for some time between 1 and 3 years.

These statistics show that today, there are half the manufactures that use AI and are equipped with the tools and algorithms yet the output isn’t very apparent. And perhaps the lacking factor is they need to embed AI deep within manufacturing from designs to operations.

About Author: Usha B. Trivedi is an engineer with Hi-Tech Engineering Services, a company providing design support services for product drafting and modeling across the globe. With years of experience in CAD modeling and rendering, she writes to provide insightful solutions for the process challenges of fabricators, engineers, architects and contractors.

Author:Usha B. Trivedi