How do electronic devices sense temperature?

How do thermoresistive sensors work, and how well do they perform?

Thermoresistive sensors work by measuring how a material's electrical resistance changes with temperature. Most materials either increase resistance when heated (positive temperature coefficient, or PTC) or decrease it (negative temperature coefficient, NTC). The key performance metric is sensitivity—how much the resistance changes per degree Celsius—and linearity, which tells you whether the relationship is predictable across a range.

A 2024 study developed a soft, stretchable thermoresistive sensor using a conductive polymer (PEDOT:PSS) embedded in elastomer. It showed a linear sensitivity of 0.46% per °C from 30 to 55°C, and when stretched by 40%, sensitivity jumped to 1.49% per °C [1]. That means if you stretch the sensor, it becomes more responsive to temperature—useful for wearable devices that move with the body. Another study used a hydrogel with ions (like barium) and found larger ions gave higher temperature sensitivity, achieving a fast response of 2.02 seconds for a 40°C temperature change and a wide range from 22 to 100°C [8]. A third group built an ultrathin fiber-mesh PTC thermistor that increased resistance by a factor of 1,000 (three orders of magnitude) within just 2°C, and it worked reliably for 400 cycles [3]. That extreme sensitivity makes it ideal for overheat protection circuits in implantable devices.

Not all thermoresistive sensors are equal. A superhydrophobic sponge-based sensor using reduced graphene oxide and polypyrrole showed a linear temperature coefficient of 0.48% per °C from 35 to 80°C, but it was designed to work underwater, where many sensors fail [11]. Meanwhile, a CMC/MXene-based sensor achieved ultra-high linearity (R² = 0.9943) and a sensitivity of -2.3575% per °C (negative coefficient), meaning its resistance dropped as temperature rose [10]. The choice of material and structure directly determines whether a sensor is best for skin contact, underwater use, or high-temperature environments.

What are the other ways to sense temperature without measuring resistance?

Two major alternatives to thermoresistive sensing are thermoelectric and triboelectric methods. Thermoelectric sensors use the Seebeck effect: when two ends of a material are at different temperatures, a voltage develops. This voltage is proportional to the temperature difference, and the material's Seebeck coefficient (in microvolts per Kelvin) tells you how sensitive it is. Triboelectric sensors generate voltage from mechanical contact and separation; when temperature changes alter the material's properties, the output voltage shifts.

A 2024 study created the first stretchable thermoelectric fiber by embedding copper(I) iodide nanoparticles in a polymer. It achieved a Seebeck coefficient of 203.6 µV/K, meaning a 10°C difference produces about 2 millivolts—enough to detect small temperature changes [7]. This fiber also detected strain and pressure simultaneously, making it a true multimodal sensor. In contrast, a 2025 study used an ionic elastomer in a triboelectric nanogenerator (iTS-TENG) and achieved a thermal sensitivity of 3.87 V/°C from room temperature to 70°C, with fast response and reproducibility over 20 cycles [5]. That's a much larger voltage signal than the thermoelectric fiber, but it requires mechanical motion to generate the signal, so it's not a passive sensor.

Another triboelectric design from 2025 used heat-resistant materials to work up to 200°C, far beyond human skin's limit of 60°C, and achieved 94% accuracy in recognizing objects by their temperature and pressure signatures [2]. This shows that triboelectric sensors excel in extreme environments where conventional electronics fail. The choice between thermoresistive, thermoelectric, and triboelectric comes down to whether you need passive sensing (thermoresistive), self-powered operation (thermoelectric or triboelectric), or extreme temperature tolerance (triboelectric with heat-resistant materials).

Can one sensor measure temperature and other things at the same time?

Yes—many modern sensors are designed to detect multiple stimuli simultaneously, such as temperature, pressure, strain, and humidity. This is critical for electronic skin, robotics, and wearable health monitors, where a single patch needs to feel touch, heat, and stretch. The challenge is to keep the signals separate so you can tell which change came from which stimulus.

A 2024 study printed a 3D electronic skin using a nanoengineered hydrogel that could detect strain, pressure, and temperature all at once. It used a combination of materials (MoS₂ nanoassemblies for conductivity, polydopamine for adhesion) and achieved precise detection of dynamic changes [4]. Another study built a bionic sensor inspired by ants and spiders, using MXene-coated foam and carbon nanotube membranes. It achieved ultra-high pressure sensitivity (986.51 kPa⁻¹) and a temperature sensitivity of 9.891 µV/K, and could map pressure and temperature distributions simultaneously using a sensor array [6]. A flexible, transparent sensor with interlocked square columns made of silver nanoparticles and PVA polymer could sense pressure, temperature, and humidity—with temperature resolution down to 0.1°C [9].

The key to decoupling signals is clever design. For example, the stretchable thermoelectric fiber [7] measures temperature via the Seebeck voltage and strain via resistance change, so the two signals don't interfere. The triboelectric sensor [2] uses an asymmetric structure that outputs two independent signals for pressure and temperature. And the hydrogel sensor [8] uses a 5×5 array to map temperature across the wrist, distinguishing blood vessels by their 0.5°C temperature difference. These examples show that multimodal sensing is not just possible—it's already being demonstrated in working prototypes.

About These Sources

This answer is built on 11 peer-reviewed studies — published from 2021 to 2026, 8 from 2024 or later, 7 in Q1 journals, collectively cited 624 times — selected as the most relevant from 15 studies that passed quality screening, drawn from 61 papers retrieved from a database of over 500 million.

Sources used in this answer

Related research questions