With the evolution of materials science and microfabrication processes, energy harvesters have become sophisticated, achieving power outputs in the range of several milliwatts, and have become a promising alternative to conventional batteries. Although their output power is insufficient to continuously operate a wireless sensor module, energy harvesters can operate small integrated circuits, including timers, watches, and ring oscillators. In this study, we emulated the human sensory nervous system to develop a battery-less sensor with a built-in analog-to-digital converter. The human sensory nervous system comprises a sophisticated sensing mechanism that digitalizes external stimuli by pulse-density modulation. To mimic this behavior, we integrated a ring oscillator with photovoltaics, allowing it to function as a luminance sensor with an event-driven operation. The oscillation frequency of the ring oscillator changes with respect to the operating voltage; hence, the output voltage of the photovoltaic modulates the frequency by more than two orders of magnitude. The sensor exhibits oscillation frequencies of 10 kHz and 7.7 MHz corresponding to luminance levels of 25–25 000 lx. Its response times are 40 μs and 15 ms when the light source is turned on and off, respectively. Battery-less sensors expand the opportunities for the application of energy harvesters in biomedical, wearable, and environmental sensing.
Receptors beneath the human epidermis detect external stimuli, including mechanical pressure, heat, and light, with high sensitivity and resolution. When such stimuli are applied to receptors, they generate serial pulses to convert the stimuli into electrical signals (Fig. 1), similar to analog/digital converters. E-skin, an artificial skin that imitates the function of human skin, has evolved in recent decades, and researchers have developed functional temperature,1–3 tactile,4,5 and luminance sensors.6–8 E-skins comprise organic and thin inorganic materials that render them flexible and stretchable for conformable attachment to curved surfaces. However, E-skins imitate the softness of human skin, and their sensing principles are the same as those of conventional electronics. Kim et al. reported a flexible, organic, artificial afferent nerve comprising pressure sensors, an organic ring oscillator (RO), and a synaptic transistor to detect pressure (1–80 kPa) and stimulate the efferent nerve.9 Wang et al. integrated emerging sensors with memristors to develop a neuromorphic sensing system.10 The RO modulates its oscillation frequency (fosc) with respect to the operating voltage, and its power consumption is lower than that of energy harvesters. Furthermore, integrated with memristors, the RO achieved spike-timing-dependent plasticity imitating in neuromorphic computing.11–13 We developed a battery-less sensor comprising lead zirconate titanate (PZT) and RO.14 The output voltage of the PZT operated the RO, which digitalized the voltage via pulse-density modulation. The system includes a built-in analog–digital converter with battery-less event-driven operations using PZT. These battery-less and event-driven operations are beneficial for implants whose batteries cannot be easily changed,15 and an artificial vision system can be an application of this system. Furthermore, we previously demonstrated normally off sensor modules that get activated by the output of energy harvesters.16 A trigger is an analog signal, owing to which the module suffers from noise, leading to an incorrect action. The proposed battery-less sensor generates AC signals and, with bandpass filters, the normally off sensor module operates when an event, which needs to be detected, takes place. Major changes in frequency further enable the development of biomimicking sensors with high sensitivity.
Circuit diagram of the light source, luminance measurement, and luminance sensor considered in this study.
Circuit diagram of the light source, luminance measurement, and luminance sensor considered in this study.
Luminance sensing with the system. (a) Frequency modulation of the ring oscillator with input voltage. (b) Luminance vs output voltage of the photovoltaics. (c) Experimental setup. Oscillation signals at (d) 100 and (e) 25 000 lx. (f) Luminance vs oscillation frequency.
Luminance sensing with the system. (a) Frequency modulation of the ring oscillator with input voltage. (b) Luminance vs output voltage of the photovoltaics. (c) Experimental setup. Oscillation signals at (d) 100 and (e) 25 000 lx. (f) Luminance vs oscillation frequency.
The response time of a sensor is an important characteristic, and we evaluated the rise and fall times of the sensor upon turning the light on and off. Figures 4(a) and 4(b) show the experimental setup in a dark room with lights on and off, respectively. Upon the switch on, the VLED increased to 5 V, and the RO started oscillating after 40 μs [Fig. 4(c)]. The quick response of the inverters and photovoltaics resulted in a fast rise time of the sensor. When the light was turned off, the oscillation halted in 15 ms [Fig. 4(d)], which was longer (by three orders of magnitude) than the time for the oscillation to start. According to the rise time, inverters and photovoltaics involve sufficiently rapid response, and Vosc decreased to 0.6 V after the switch was off. Such a slow response can be ascribed to the residual charge at the gate capacitors of the inverters and photovoltaics. The removal of the charge can yield a fast fall time. To demonstrate luminance sensing, we blocked the light exposed to the sensor manually [Fig. 4(f) and Movie S1 (supplementary material)]. The sensor exhibited high fosc when exposed to light; upon blocking the light path with the hand, fosc decreased. Thus, the system functioned as a luminance sensor.
Evaluation of the response of the sensor. The sensor with a light in (a) on and (b) off modes. Transient behaviors of oscillation upon turning (c) on and (d) off the light. Change in the oscillation frequency of the sensor (e) in a dark room and (f) exposed to light.
Evaluation of the response of the sensor. The sensor with a light in (a) on and (b) off modes. Transient behaviors of oscillation upon turning (c) on and (d) off the light. Change in the oscillation frequency of the sensor (e) in a dark room and (f) exposed to light.
We developed a battery-less sensor that imitates the human sensory nervous system. The oscillating frequency of the RO changes with respect to the operating voltage; hence, the output voltage of the photovoltaics modulates the oscillating frequency by more than two orders of magnitude. The sensor detected luminance ranging from 25 to 25 000 lx with a dynamic range of 58 dB. Its response times were 44 μs and 14 ms upon turning on and off the LED, respectively. Battery-less sensors can help expand the opportunities for the application of energy harvesters in biomedical, wearable, and environmental sensing.
SUPPLEMENTARY MATERIAL
See the supplementary material for demonstration of the biomimicking sensor under illumination (Movie S1).
This work was supported in part by JST CREST (Grant No. JPMJCR15Q4) and JSPS (Grant No. 22K14213), Japan.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Shunsuke Yamada: Conceptualization (lead); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (lead); Software (lead); Validation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (equal). Hiroshi Toshiyoshi: Funding acquisition (lead); Project administration (lead); Supervision (lead); Writing – review & editing (equal).
DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.
