The sensing-decision-action loop
Connect sensors, control logic and an output in one simple model.
Sense a condition
A sensor provides information about temperature, position, light or another condition. Its range, accuracy and update rate affect what a system can observe.
Make a decision
Control logic interprets a signal and chooses an action. A threshold and a complex model both need clear assumptions about normal and unusual conditions.
Act and observe again
An output may switch a device, move a mechanism or notify a person. Feedback measures what happened after an action, allowing the next decision to use updated information.
Plan human oversight
Sensor errors, failed connections and unexpected conditions need deliberate handling. Intelligent features do not remove the need to define responsibility, verify behaviour and provide a way to intervene.
Take the next step
Write down one question this guide raised. Identify what you would need to observe, compare or learn before turning it into a project.
Explore this fieldFurther reading: NASA Science: technology and autonomous systems. This is an original introductory overview; the linked resource offers wider context. Tanevor is independent of these organisations.