Can a sensor that gives accurate readings on a laboratory bench be trusted on a moving person? Not automatically. A sensor detects an input from its physical environment, but placement, motion, interference, software, and the user’s behavior can all change what it measures. Testing in the setting where a device is meant to work asks whether the full system still produces reliable, useful information when those conditions are present. [1]
A sensor measures a specific input
A sensor detects and responds to a physical input, such as temperature, pressure, light, flow, or a chemical signal. In a medical setting, it may report a biological, chemical, or physical process to a user or health professional. [1] The reading is not the process itself; it is an electrical or digital representation produced by a chain of sensing, calibration, and interpretation.
Consider a wearable patch intended to detect breathing sounds. On a quiet bench, a microphone-like sensor may register a clean sound. On a moving person, clothing friction, body movement, room noise, and imperfect contact can become signals too. The device must distinguish the target from these competing inputs. A strong bench result answers whether the sensor can respond under controlled conditions, not whether it can do so reliably in daily use.
The body and environment alter signals
When a sensor touches skin or sits near the body, its interface matters. Sweat, temperature, pressure, movement, and the shape of the body can change contact and signal quality. The surrounding environment adds further variation: sound reflections, bright light, electromagnetic interference, or changes in activity can affect some kinds of sensors. These are not minor details if the device is intended to work through them.
NIBIB describes wearable technologies that track measures such as heart rate and sleep patterns, and pulse oximeters that use changes in light absorption to estimate heart rate and blood oxygen. [1] Those examples use different physical signals, so the relevant tests differ. A light-based sensor needs evaluation under varying skin contact and light conditions; an acoustic sensor needs evaluation amid real sounds and movement. Testing should follow the measurement mechanism.
A measurement must support its intended job
Accuracy is not a single property divorced from use. A device may be designed to monitor, modulate, or supplement bodily functions. [1] A monitor that flags a possible change has a different job from a device that controls another device. For instance, NIBIB describes a continuous glucose monitor that can signal an insulin pump to adjust delivery. The linked system makes dependable sensing and communication especially important, because a measurement can influence an action. [1]
The FDA describes multiple regulatory pathways for medical devices, including premarket review requirements for many device types and clinical studies that can collect safety and effectiveness data. [2] That framework reflects a broader point: evidence has to fit the claim. A laboratory demonstration may support that a sensor responds to a signal. Evidence from people in the intended setting is needed to understand performance amid actual use, and a clinical benefit claim requires support appropriate to that claim.
Test the whole chain, not just the component
A useful evaluation follows the signal from input to result: does the sensor detect the intended physical change, does the electronics preserve it, does software interpret it correctly, and can the user or clinician act on the report? Each link can fail in a different way. A false alarm may arise from noise; a missed change may arise from poor contact; a correct measurement may still be unhelpful if it arrives too late or is misunderstood.
The FDA describes device oversight that includes labeling, quality systems, and reporting certain serious device incidents, alongside pathways for marketing and clinical investigation. [2] These are parts of a larger lifecycle in which performance matters beyond the bench. Testing where a device will actually be used reveals whether its physical measurement survives the ordinary disturbances of that environment and whether the output remains meaningful for its intended purpose.
Explore the physics: an explanatory diagram
Sources & further reading
Prepared for the independent PSI Center reading library. About edition dates and source checking.
This explainer describes scientific principles. It does not assess individual treatment choices.
