Despite all the promise of digital transformation and the role of artificial intelligence in powering the factories of the future, its adoption is still relatively nascent in much of the manufacturing sector. About the Author Ted Plummer, senior product manager and resident artificial intelligence expert at industrial 3D printing company Markforged. There are several reasons for this, including a lack of understanding of what AI really is and what changes it will bring. Separating fact from (science) fiction can be challenging. Confusion, associated with uncertainty, engenders des craintes et des idees fausses, qu'il s'agisse de risques pour la sécurité, de pertes d'emplois, de parte de contrôle et de ce que la technologie peut et ne peut pas do.
Myth 1: AI is the end goal
There is a misconception that AI itself is a benefit: I have had countless conversations with clients who have misunderstood that AI is a mechanism, not a benefit. I've heard "I'll wait for him to 'AI'" more times than I can count. The reality is that the benefit of AI is not the process itself, but like any type of data analysis, the value of AI comes from its ability to solve problems faster, speeding up production. AI is the how, not the why. The second part of the AI equation is federated learning. Apple or Android smartphones use federated learning technology to improve with each typed text message based on how individual and group users interact with their keyboards. Similarly, our network of more than 10,000 securely connected 3D printers applies this AI technology to enable every machine to 'get smarter' with every print, while maintaining the highest standards of privacy, integrity and confidentiality of customer data. By analyzing data from the "fleet" of printers, the AI can detect corrections or adjustments made on a regular basis, such as when throw angles or fill patterns aren't quite right. These improvement opportunities can then be fed back into the system, improving the collective performance of printers without human intervention.
Myth 2: AI is insecure and based on proprietary data
There is a misconception that since AI is data-driven, it forces those who use it to share their intellectual property (IP) for profit. This is not the case. When it comes to AI in 3D printing, the customer's IP address and part data remain separate within safe limits. It is not this proprietary information that drives the federated learning described above, but anonymous metadata. It is the information that is essentially collected into a data "set" that allows machines to learn and improve. It is not possible to recreate any of the source IP addresses from the collective data. However, as with any data-driven technology, security is always of the utmost importance when it comes to using AI. Making sure you're based on a secure platform with the integrity and confidentiality of customer data is critical, and an ISO 27001 certification is a great way to show you've invested in risk management.
Myth 3: AI is constantly evolving, making its results unpredictable and unsuitable for repeatability.
For highly regulated industries like aerospace, repeatability is paramount. When creating parts for airplanes, for example, the 10.000th printed part should be exactly the same as the first. For this reason, AI, and federated learning in particular, is often shunned by regulated industries. Its benefits of progressive learning and enhancements are considered to be in contradiction with strict and critical lifetime safety requirements. However, industries like aerospace, where repeatability is required, can still benefit from AI-based technologies. It can be used for design iterations, for example to help refine and perfect aircraft parts in the early stages of development. Once the team is happy with the room setup, the system can be 'locked down' to ensure no further changes or fleet data updates are incorporated. At this point, the technology can be used as a verification tool to ensure that there are no bypasses in the printing process and that each part is exactly the same as the last. In the long term, the same technology can provide even greater repeatability by detecting and compensating for changes in system behavior, such as lack of lubrication or machine wear.
Myth 4: AI will replace humans and take away our jobs
This myth still resides a lot in the world of science fiction. I say let the machine take over when it comes to machine problems! Very few operators, engineers or industrial designers would complain if machines could "repair themselves", freeing them from mundane troubleshooting tasks and allowing them to carry out their daily tasks. Instead of making us lazy or redundant, the use of artificial intelligence and machine learning helps drive innovation and smarter work practices. In manufacturing or product design, instead of focusing on more process-oriented "what" and "how" questions, it allows engineers to ask "why" and "what if" questions and explore the implications of different scenarios when it comes to increasing efficiencies or creating new products, which ultimately leads to greater business opportunities.
Myth 5: the cost of AI is holding back its adoption
There are two common responses I hear when talking to customers about our AI-powered machines: (1) "I can't believe how affordable this is!" Or (2) "It costs too much!" As with any technology in development, there are those who can see the value it can bring and those who see it as an expensive luxury. We are beginning to see this shift as AI moves beyond the early adoption phase. Those who advocate AI-based solutions in the factory are focusing on the value they can deliver, essentially enabling machines to solve machine-related problems, freeing up engineers and operators to invest their efforts in innovation, product development, and more. products and other people-related endeavors. . It's important to remember that many of these myths exist because not all AIs are created equal. To be an effective tool, AI requires access to vast amounts of data; machines cannot "learn" without a constant stream of reliable data. Before you invest in any AI-powered technology, make sure you have a reliable data source that can scale with the machines you power.