Robotics & AI Vocabulary

robot — a physical machine capable of performing actions in the physical world, often programmed to complete specific tasks: “The robot moved boxes from one shelf to another.”

robotics — the field concerned with designing, building, and operating robots: “She studied robotics at university.”

industrial robot — a robot designed for tasks such as manufacturing, assembly, or packaging, typically used in factories or similar settings: “The industrial robot welded each part with the same precision every time.”

automated / autonomous — automated describes a system that performs a task automatically according to programmed processes or triggers; autonomous describes a system able to make some decisions or adapt its actions without constant direct human control, though autonomy doesn’t mean unlimited independence: “The assembly line was fully automated, but the delivery robot was autonomous enough to choose its own route.”

robotics / automation — robotics is the field concerned with designing, building, and operating robots; automation is the use of systems or processes to perform tasks automatically, and automation doesn’t necessarily involve robots at all. Robotics also doesn’t mean every robot is fully automated or autonomous: “The plant used automation to sort packages with no robots involved, while a separate robotics team built a prototype that still needed a person to operate it.”

sensor — a device that detects information from the environment, such as light, distance, or temperature: “A sensor stopped the robot before it collided with the shelf.”

actuator — a component that produces physical movement in a machine, converting a signal into action: “The actuator moved the robotic arm into position.”

machine vision / computer vision — technology that allows a system to interpret images or visual information: “Machine vision let the robot identify defective parts on the line.”

human-robot interaction — the ways in which people and robots communicate, cooperate, or work alongside one another: “Researchers study human-robot interaction to make workplaces safer.”

repetitive task — an action performed the same way many times, often a good candidate for automation: “Sorting identical parts is a repetitive task well suited to a robot.”

robot / AI — a robot is a physical machine capable of performing actions in the physical world; AI is computational technology, not a physical object. A robot may use AI to help it sense or decide, but not every robot uses AI, and AI does not require a physical robot at all: “The robot on the factory floor followed simple programmed steps with no AI involved, while a separate AI system analyzed sales data from an office computer.”

sensor / actuator — a sensor detects information from the environment or system, such as light, distance, or temperature; an actuator produces physical movement or action, converting a signal into a physical effect. In short, a sensor takes information in, and an actuator makes something happen: “A sensor detected the obstacle, and an actuator adjusted the arm’s position in response.”

machine vision / computer vision — both terms describe a system’s ability to interpret images or visual information, and their use overlaps considerably; machine vision is often used for practical, industrial applications such as inspection, measurement, or guiding equipment, while computer vision is the broader term and can describe visual interpretation in many kinds of systems, not only industrial ones. The exact usage can vary by industry or context: “The factory’s machine vision system checked each part for defects, while a separate research team applied computer vision techniques outside any industrial setting.”

artificial intelligence — computational systems capable of tasks such as prediction, classification, pattern recognition, generation, or interpretation: “Artificial intelligence is used in fields ranging from healthcare to logistics.”

AI model — a computational system trained to identify patterns and produce predictions, classifications, decisions, or generated outputs, depending on its design: “The company built an AI model to predict equipment failures.”

machine learning — a way of developing AI systems using data, so that patterns can be learned rather than every rule being individually programmed: “Machine learning allowed the system to improve as it processed more examples.”

training data — the data used during the process of developing or training a model; not every AI system is trained the same way: “The model’s accuracy depended heavily on the quality of its training data.”

automation / AI — automation follows rules or processes to perform tasks automatically; AI refers to computational systems capable of tasks such as prediction, classification, pattern recognition, generation, or interpretation. A system may combine both, but not all automation is AI: “The assembly line used simple automation, while a separate AI system predicted which machines were likely to need maintenance.”

generative AI — a category of AI that produces new content, such as text, images, audio, video, or code, depending on the system: “She used generative AI to draft the first version of the report.”

prompt — an instruction, question, or other input given to an AI system to guide its response or output: “He wrote a detailed prompt describing exactly what he needed.”

output — what a system produces after processing input: “She reviewed the output before using it in her presentation.”

generative AI / AI — generative AI is one category of AI, specifically the kind that produces new content; AI is the broader term covering many other tasks as well, such as prediction and classification, so the two terms aren’t synonyms: “The company used generative AI to draft marketing copy and a separate, non-generative AI system to detect fraud.”

chatbot / AI assistant — a chatbot is software designed primarily for conversation through text or voice; an AI assistant may also use conversation but can be designed to help perform a wider range of tasks. In everyday use, the terms can overlap: “The chatbot answered simple questions, while the AI assistant could also schedule appointments and send reminders.”

prompt / output — a prompt is the instruction, question, or other input a person gives to an AI system; the output is what the system produces after processing that input. In short, a prompt goes in, and an output comes out: “She refined her prompt until the output matched what she needed.”

hallucination — an output generated by an AI system that presents incorrect, unsupported, or invented information as though it were reliable; not every AI error counts as a hallucination: “The report contained a hallucination — a citation for a source that didn’t exist.”

verify — to check that something is true, accurate, or correct: “She verified the figures before including them in the summary.”

human oversight — the practice of having a person review, check, or approve the work of an automated or AI system: “Human oversight caught the error before the report was sent out.”

plausible / accurate — a plausible response sounds reasonable and believable; an accurate response is actually correct. An AI response can sound plausible without being accurate, which is why verification matters: “The answer sounded plausible, but it wasn’t accurate once she checked it against a reliable source.”