This is because the engineers developing them are largely ignoring the fundamental realities of plant biology and farming environments.
Researchers at Michigan State University, Sejong University, Seoul National University and Wageningen University wrote a review paper on the subject.
They claimed that the “skyrocketing” development of these small, flexible sensors was heavily driven by engineering advancements, but was simultaneously plagued by a glaring lack of proper testing on plants in real-world settings.
The authors warned that this communication gap between WPS developers and the plant research community was hindering the practical adoption of this technology in the agricultural sectors that needed it most: open fields, greenhouses and vertical farms.
Wearable plant sensors are designed to be attached directly to a crop, measuring vital internal and external signals related to growth, water status, nutrients, stress and microclimate.
Proponents believe these devices are key to the “internet of plants”, which could drastically improve resource efficiency, reduce crop losses, and automate precision agriculture.
Lab testing is not good enough
The review’s core finding was a critique of current experimental practices.
The researchers analysed 93 experimental studies on WPSs and noted a troubling trend of inadequate in-plant testing.
They found that only 15 out of 93 experimental studies — a mere 16 per cent — actually tested their sensors in the field, and that even these few field tests were typically very short, with a median duration of just 48 hours.
The problem runs deeper than a lack of field testing, however. The researchers found that most studies fail to meet the most basic standards of plant science, with environmental blind spots, poor replication and a lack of cross-validation emerging as key issues.
For instance, a massive 88 per cent of all studies did not adequately report the environmental conditions (such as light intensity, temperature or humidity) in which the plants were grown or tested.
This lack of proper reporting suggested an alarming lack of awareness that plants are complex organisms that respond to their environment.
At the same time, approximately 73 per cent of studies tested the sensor on only one plant, or failed to clearly report the number of biological replicates used.
This ignores natural biological variation and makes drawing reliable conclusions about a sensor’s performance virtually impossible.
Finally, in 68per cent of studies, the data measured by the WPSs on the living plant was not cross-validated against an independent, standard measurement. This means developers often cannot prove the signal being recorded is physiologically meaningful.
The study’s authors stressed the importance of proper testing, saying it could not be assumed that the calibration between phenomenon and signal would hold once a WPS was attached to a plant, as the signals generated by a plant were often noisier than those in vitro.
What developers need to know: Key signals and pitfalls
The authors highlighted a wide array of signals WPSs could measure, which are critical for different aspects of crop management. These included environmental factors like light, vapour pressure deficit (VPD) and pesticides, as well as factors related to plant physiology, such as water status, nutrients, phytohormones and volatile organic compounds (VOC).
For example, many studies incorrectly use the unit lux (focused on human eye sensitivity) instead of the correct plant unit, photosynthetic photon flux density (PPFD). The intensity and spectrum of light are non-negotiable drivers of plant growth and must be controlled and reported correctly.
VPD is a crucial indicator of the plant’s water stress and the driving force of transpiration; measuring it alongside leaf temperature and relative humidity can provide early warnings of drought.
Furthermore, while several sensors can detect residues on surfaces, developers must conduct physiologically relevant testing to assess real-world performance amid plant surface components, humidity, and other environmental factors.
In terms of to plant physiology, measurements of leaf surface humidity (LSH) and sap flow are vital for optimising irrigation. However, developers must consider the impact of sensors that block stomata or alter long-term physiological responses.
At the same time, new sensors can track the movement of key ions like nitrate and potassium within a plant, offering huge potential for precision fertilisation management.
Tracking hormones like abscisic acid (ABA), salicylic acid (SA) and ethylene is also a promising avenue for early stress detection. Ethylene, a gaseous hormone produced under stress, has been shown to increase within just one day of water stress in some crops.
Finally, VOCs emitted by plants are an early warning system for stress. They can differentiate between mechanical damage and biotic attacks like fungal infection, sometimes within just four days.
The environment determines the design
The authors stress that developers must decide from the outset which agricultural system their sensor is intended for, as each environment presents unique challenges. In the open field, WPSs must endure direct sunlight, UV radiation, rain, wind, and wide temperature fluctuations.
Robust encapsulation, flexible mounting, and adhesive backing layers are essential to prevent the sensor from degrading or falling off due to mechanical stress. Additionally, low-power electronics, efficient communication protocols (like LoRa for long-range), and solar or kinetic energy harvesting are critical for continuous, long-term monitoring.
Early detection of abiotic stresses (drought, nutrient deficiency) and diseases is highly valuable. Future systems could integrate WPS data wirelessly with drones and field robots for autonomous response.
In greenhouses, despite their more controlled environments, sensors must cope with high humidity, potential condensation, and high temperatures. High humidity requires waterproof or water-resistant encapsulation to protect sensitive electronic components and ensure reliable readings over time. As such, WPSs can be integrated into the greenhouse management system to create a dynamic feedback loop, allowing growers to adjust climate set points based on real-time plant feedback.
When it comes to vertical farming systems (VSFs), precision control is key, as VFSs simplify durability but introduce new complexities related to light. The specific LED lighting used for plant growth can introduce interference and signal noise into optical sensor measurements. Sensors may therefore need synchronization with the lighting cycles or optical filtering.
As VFSs are energy-intensive, integrating WPS data directly into machine learning models could predict plant growth and resource requirements. This would allow growers to optimise nutrient recipes and lighting schedules in real time, ultimately reducing operating costs.
The path to adoption
To bridge the gap between lab and farm, the review proposed a clear protocol for developers. They must use an adequate number of biological replicates (at least three per treatment), and confirm sensor readings in vivo using a completely independent standard measurement. They must also test for long-term stability and whether the sensor itself negatively impacts plant growth.
The authors concluded: “Although outside the scope of WPSs as defined in this review, other sensing approaches, such as RGB imaging, hyperspectral imaging, Raman, and infrared sensing, share similar goals of non-invasive, real-time monitoring.
“These technologies may serve as complementary tools that, when combined with WPSs, could further enhance plant phenotyping and crop monitoring in the future. Altogether, we propose that WPSs can be a very valuable addition to crop monitoring if their testing on plants is done properly.”
Source: ACS Sensors Let’s Get Real: Are Wearable Plant Sensors Ready for Crop Monitoring?” https://doi.org/10.1021/acssensors.5c02510 Authors: Hoh Donghee, et al.
