skip to main content

Smart Manufacturing

According to the National Institute of Standards and Technology (NIST), Smart Manufacturing is defined by a "fully-integrated, collaborative manufacturing systems that respond in real time to meet changing demands and conditions in the factory, in the supply network, and in customer needs." If pursued correctly, smart manufacturing can lead to higher quality products, improvements in productivity, and increases in safety.

One aspect of Smart Manufacturing focuses on capturing, gathering and analyzing data to make plant management decisions based on facts. To do so efficiently requires the synchronization of data-capturing devices and machines on the plant floor with an information system where data can be curated and analyzed. New advances in the Internet of Things for industrial facilities, artificial intelligence and cyber security can be considered as well as developing new tools to intelligently monitor equipment and overall performance conditions. Predictive maintenance through the development of new Artificial Intelligence and Machine Learning algorithms can be applied to great impact.

Another key aspect of Smart Manufacturing is the introduction of increased automation and flexibility through new collaborative and learning techniques for robots on the plant floor. Traditional industrial robots on the plant floor are programmed to complete a single task at a time. The robot must be reprogrammed if it is to undertake another task. Advances in perception and artificial intelligence provide for the possibility of robots being able to learn how to complete different tasks by sensing their environment and adjusting accordingly.

The future of industrial robotics may well be represented by the 2020 National Science Foundation solicitation entitled, "National Robotics Initiative 2.0: Ubiquitous Collaborative Robots." That solicitation called for the development of collaborative robots (co-robots) and defined a co-robot as, "a robot whose main purpose is to work with people or other robots to accomplish a goal. An ideal co-robot is an adaptable partner, not limited to a narrow set of specified interactions or functions, but able to significantly enhance team performance despite changes in its role, its teammates, or the team's collective goals." Challenges to the wide adoption of such robots include, to name but a few, designing controls for facilitating ubiquitous interaction and for making co-robots inherently safe, incorporating learning efficiently from direct experience with people or other robots, and enabling robotic systems to reliably perceive, act, plan, and learn.

Closely linked to the emergence of intelligent robots is the notion of digital twins. A digital twin creates the virtual model of an asset, process, or system by using the data obtained from sensors in the systems or asset. Through the development of algorithms, digital twins can be used to speed up reprogramming of robots through simulation potentially reducing the time required and the elimination of unplanned downtime. In this regard, the incorporation of IoT and cloud platforms in the manufacturing plants are key enablers of digital twins.