AI in Manufacturing Overview 115

acoustics

Relating to sound. Acoustic monitoring analyzes audible and inaudible process noises.

actuators

A device that translates a digital signal into physical action. Actuators are located in robots and industrial machines.

AI

AI. A computer program that uses algorithms to enable a machine or computer to imitate intelligent human behavior. Artificial intelligence allows machines to perform a process with autonomy.

AIoT

Artificial Intelligence of Things. A combination of machine learning algorithms and a network of physical devices with embedded computing systems that can send and receive data. AIoT allows devices to exchange data and make predictions about a specific dataset.

algorithms

A logical and mathematical expression that models a process or action. Algorithms are coded into a computer program that forms the rules by which a machine learning model will process data inputs and produce outputs.

application domains

Devices that determine how software acts externally with the environment. Application domains for artificial intelligence (AI) can include robotics or natural language processing among many others.

artificial intelligence

AI. A computer program that uses algorithms to enable a machine or computer to imitate intelligent human behavior. Artificial intelligence allows machines to perform a process with autonomy.

Artificial Intelligence of Things

AIoT. A combination of machine learning algorithms and a network of physical devices with embedded computing systems that can send and receive data. The Artificial Intelligence of Things allows devices to exchange data and make predictions about a specific dataset.

augmented reality

AR. A technology that superimposes a computer-generated image on a view of the real world. Augmented reality methods can generate real-time data and visually alert the wearer to potential hazards.

CNC

Computer numerical control. A combination of software and hardware that directs the operation of a machine. CNC machines are used for many manufacturing operations, such as cutting and grinding metal parts.

cobot

A robot designed to interact with humans in a shared workspace. A cobot, or collaborative robot, can be programmed to learn a specific task in order to assist humans.

computer numerical control

CNC. A combination of software and hardware that directs the operation of a machine. Computer numerical control machines are used for many manufacturing operations, such as cutting and grinding metal parts.

computer vision

A deep learning capability that enables an artificial intelligence (AI) to detect and identify objects by recognizing visual patterns through a camera or on a screen. Computer vision can be useful for supervised or unsupervised machine learning but requires very large amounts of visual data.

Coordinate measuring machines

CMM. A sophisticated measuring instrument that uses a suspended probe to measure parts in three-dimensional space. Coordinate measuring machines operate using either contact or noncontact methods.

data

A collection of numbers, information, or facts that is used as a basis for making conclusions. Data collected by manufacturers can include records of machine issues, customer order history, and raw material inventory.

data science

A field of study that uses statistics and data analysis with computing technology. Data science allows manufacturers to gain valuable insight from historical data to improve operations.

data scientist

An individual who uses computing systems to perform data analysis in order to interpret and present insights gained from data. Data scientists can help manufacturers operate more efficiently by adjusting processes based on data analysis.

deep learning

An advanced form of machine learning that uses neural networks with multiple hidden layers. Deep learning algorithms can enable machines to exhibit advanced, human-like behaviors but require significant amounts of data.

demand forecasting

The process of analyzing historical manufacturing data to uncover trends used to optimize future decisions. Demand forecasting can be enhanced by deep learning models that can potentially manage manufacturing usage efficiency autonomously.

deviations

The difference between a standard and a result. Deviations can lead to parts that do not meet specifications.

die

A tooling component of stamping presses that attaches to the slide and bolster. Dies perform various stamping operations including cutting and forming metal parts.

edge computing

The practice of running software and processing and storing data on multiple local devices located at the data sources. Edge computing distributes processing tasks across multiple edge devices.

G code

A method of programming that pairs address letters with numerical values to form words. G code programs are used in additive manufacturing and CNC machining.

GenAIoT

Generative Artificial Intelligence of Things. A combination of deep learning algorithms and a network of physical devices with embedded computing systems that can send and receive data. GenAIoT creates new data rather than making a prediction about a specific dataset.

Generative Artificial Intelligence of Things

GenAIoT. A combination of deep learning algorithms and a network of physical devices with embedded computing systems that can send and receive data. Generative artificial intelligence of things creates new data rather than making a prediction about a specific dataset.

hallucinations

A software output that creates false or misleading information due to faulty or incorrect input data. Hallucinations are often associated with inaccurate Generative Artificial Intelligence (GenAI) responses presented as fact.

human-robot collaboration

HRC. A manufacturing process where a human operator and a robot work closely together to complete a task. Human-robot collaboration allows for more flexibility in the use of robots while improving the efficiency of operations.

Industrial Internet of Things

IIoT. A network of physical devices used in manufacturing containing embedded computing systems that allow them to send and receive data. The Industrial Internet of Things allows devices to exchange data and automate processes without any human intervention.

logistics

Planning, managing, and executing the movement of materials and products. Logistics involves physically moving goods between locations as well as determining the routes, schedules, and methods used to move them.

machine events

An action, change or alarm that occurs on a machine. Machine events can be monitored from a controller or remotely through connected computer, smartphone, and tablet software applications.

machine learning

ML. The process that enables a digital system to analyze data in order to build predictive models. Machine learning systematically solves problems using highly complex algorithms.

metal forming

A manufacturing process that forms parts by shearing, stretching, bending, or compressing sheet metal using punches and dies. Metal forming includes hydroforming and stamping.

ML

Machine learning. The process that enables a digital system to analyze data in order to build predictive models. ML systematically solves problems using highly complex algorithms.

model drift

Changes or degradation of data used with artificial intelligence systems. Model drift is a common reason why machine learning systems require retraining.

model training

The use of specially designed data to enable artificial intelligence (AI) systems to recognize patterns. Model training is used in AI to make predictions and complete tasks in a way that mimics human thinking.

natural language processing

NLP. A deep learning capability that enables an artificial intelligence (AI) to detect and identify language patterns, audio speech, or written text. Natural language processing requires very large amounts of data.

neural networks

A web of connected nodes designed to resemble the structure and function of neurons in the human brain. Neural networks build advanced machine learning models for artificial intelligence (AI) in robots and other machines.

predictive maintenance

PdM. A maintenance approach that involves collecting data related to machine operation in order to service a machine before maintenance issues arise. A predictive maintenance approach involves performing maintenance on a scheduled basis and helps prevent unscheduled downtime.

prescriptive maintenance

RxM. A maintenance approach that uses artificial intelligence and machine learning to report potential outcomes of operating conditions, recommend actions, and suggest repairs. Prescriptive maintenance is emerging as a possible improvement on predictive maintenance.

quality

An approach to manufacturing that focuses on customer satisfaction. Quality products conform to specifications, are free of defects, and meet the requirements of their anticipated use.

radio frequency identifiers

RFID. A technology that uses electromagnetic fields to automatically identify and track tags attached to objects. RFID tags consist of a tiny radio transponder, radio receiver, and radio transmitter.

reinforcement learning

A process in which an artificial intelligence (AI) learns the best actions to achieve a goal through trial and error. Reinforcement learning enables an AI to learn details about inputs and produce correct outputs autonomously.

robotics

A field of science that is focused on programmable mechanical devices. Robotics enable the work of a person to be done by a robot with a higher degree of accuracy.

robots

A mechanical device that can be programmed to perform a variety of complicated, repetitive tasks. Robots are used to automate manufacturing processes.

semi-supervised learning

A process in which machines use both human labeled data and unlabeled data to define desired outputs. Semi-supervised learning uses a combination of supervised and unsupervised learning.

sensors

A device that detects a change in a physical stimulus and turns it into a signal that can be measured or recorded. Sensors may be connected to a machine or system in order to collect operational data that is later analyzed.

smart manufacturing

The information-driven, event-driven, efficient, and collaborative orchestration of business processes along with physical and digital processes within plants, factories, and across the entire value chain. Smart manufacturing uses the Industrial Internet of Things to connect devices and operations.

stamping presses

A machine tool that moves up and down continuously to cut and shape sheet metal using a die. Stamping presses often require dual palm buttons for starting.

supervised learning

A process in which a human operator labels data inputs for a machine learning model and defines the desired outputs the model should produce. Supervised learning typically requires extensive human labor in order to prepare and label datasets for the machine.

supply chain

A complex network of companies and suppliers that produce and distribute a product. A supply chain consists of a company, its suppliers, its distributors, and its customers.

tiny machine learning

TinyML. Describes the process of using machine learning on small devices like sensors or microcontrollers. Tiny machine learning processes can run on low-powered devices and can reduce dependence on cloud servers.

unsupervised learning

A process in which a machine learning model processes and analyzes unlabeled data and produces outputs without human interaction. Unsupervised learning models help machines discover patterns in data that humans may not recognize.

work envelope

The defined area of space through which a robot can move. The work envelope, also known as the work cell, is dangerous for operators to enter unless the robot is powered down.