Could this be the time? Ever since I started my first job as a design engineer, I’ve heard it phrased in one of many ways, but the basic premise was always the same. We are going paperless in the factory. No more 2D drawings, shop traveler sheets or written step-by-step assembly instructions. Everything from design intent, manufacturing dimensions, and quality criteria will be captured and communicated downstream by the 3D model.
It has almost been 20 years since I first heard those words. Investing in paper and pulp industry over that same period would not have been a bad investment. Sure, we do use more tablets and smartphones for some of these tasks, but nowhere to the extent most claims have asserted. Could the 20th time be the charm? There are reasons to be hopeful. Welcome cognitive assistants.
Cognitive assistants or virtual assistants, as they are frequently called, are becoming mainstream. Most of us have seen one of the ‘Hey Google’ commercials or have summoned Alexa to tell us the time and local weather forecast or turn on the living room lights. We can even order just about anything from Amazon using simple voice commands on Alexa. While some of these tasks appear simplistic, they do hold great promise to transform design and manufacturing, as we experience it today. The underlying technology powering these devices is artificial intelligence (AI), or more specifically a combination of natural language processing (NLP) and machine learning. The true power in these cognitive assistants is their ability to both retrieve the data and do it in a conversational manner that is seamless with the way we humans like to communicate namely by voice, gestures, and body language.
What if the NLP technology could be extended further to allow assemblers and maintenance technicians to query 3D models in lieu of reading step by step assembly instructions or dig through reams of papers to find repair procedures? By using simple voice commands, the assembler could interact with the CAD model, rotating and exploding the individual components for greater insight and visibility. The cognitive assistant would augment an assembler effort by sounding off the next step in the build sequence. A cognitive assistant ability to bi-directionally interface with the real world is further enhanced if it was equipped with vision and perception capability. Outfitted with computer vision, the cognitive assistant would be capable of performing real-time quality inspection while the assembler is carrying out each build step. For example, it would not be too difficult to train the cognitive assistant to detect defects or variations in the process or identify the number and types of bolts installed and whether the right washer and threadlocker compound was used. Computers are much better than humans in detecting such variations.
On the design side, what if cognitive assistants were outfitted with conversational language capabilities and are able to translate spoken words into 3D models? Combining simple hand gestures and voice commands, anyone would be able to naturally create part models on the fly. How transformative and democratizing would that be? The freedom and ability to design and create parts will be available to millions with no formal training in traditional CAD software and toolsets. To physically realize those designs, one would simply use simple voice commands to push the models to a nearby 3D printer.
Today, traditional workflow in product development usually starts with paper sketch that is translated to CAD model by trained CAD designers. What if each of us can readily translate our thoughts and product concepts into their physical manifestation through the use of capabilities each of us already possess: voice and hand gestures? This has the potential to close the widening gap between disadvantaged producers who might otherwise not have access to extensive toolsets.
Cognitive assistants with conversational NLP capabilities and machine vision hold great promise for transforming design and manufacturing. The last few decades we saw advances in computer aided design (CAD), and computer aided manufacturing (CAM). The time has come to replace the word “computer” with “cognitive.” Computers are great but have limitations in the way we interface with them mainly through a mouse and keyboard. Cognitive on the other hand opens up our ability to leverage computer intelligence and software applications by integrating voice, vision, and hand gestures to transform the way we conceptualize and realize products.