How Advanced Sensing is Transforming Agriculture

Spectroscopy In Agriculture
Grating-Based Spectrometers
Solving the Value Problem with Shortwave Infrared
Fourier Transform Spectroscopy
The Fabry-Pérot Interferometer (MEMS-FPI)
Gas Sensing in the Mid-Infrared
Conclusion
Want to Learn More? Watch the Full Webinar.
Related products

Precision agriculture demands faster, nondestructive insight into soil, water, and crop health. Miniaturized and inline spectroscopic technologies are supporting these needs by enabling laboratory-grade analysis in the field.

Introduction

Agriculture is turning to optical sensing to make faster, better-informed decisions. This shift is described as smart agriculture, digital farming, or agri-photonics. At the center of these applications is spectroscopy, which is the analysis of how materials absorb and reflect light at different wavelengths to reveal properties invisible to the eye.

Mobile Padding

Gaetano Panagia, Technology and Market Expert

Gaetano Panagia, Technology and Market Expert

Responsive Padding

Spectroscopy In Agriculture

Spectroscopy serves two broad purposes in agriculture: characterizing the external color of a product and determining its internal composition. When broadband light is directed at a sample such as an apple, part of the light is absorbed and part of it is reflected. Analyzing the reflected light, particularly in the near-infrared (NIR) spectrum, can distinguish among apple varieties such as Golden Delicious, Granny Smith, and Star King Delicious, revealing compositional details that appearance alone cannot. This paper traces several spectroscopic principles, the agricultural applications they enable, and the development pathways that can help translate photonic sensing technologies into practical agricultural solutions.

Grating-Based Spectrometers

The most widespread spectroscopic approach uses grating-based instruments. Light reflected from the sample enters the instrument through a slit, is collimated, and reaches a grating that separates broadband light into its individual wavelengths. Each wavelength is then focused onto a distinct pixel of a linear image sensor, allowing the intensity at every wavelength to be measured.

The value of this approach in digital farming depends on the wavelength range under investigation. In the ultraviolet (UV) range, between roughly 200 nm and 400 nm, analysis can detect water pollutants including nitrates and nitrites, supporting environmental monitoring and regulatory compliance. In the visible and NIR range, roughly 350 nm to 850 nm, silicon sensors can detect chlorophyll absorption, which is a strong indicator of crop health. A compact spectrometer mounted on a drone can survey a field and flag areas needing attention by reading the visible spectrum. In the shortwave infrared (SWIR), from about 800 nm to 2,500 nm, lies perhaps the most crucial range for digital farming. Remote soil analysis of moisture levels and fertilizer needs support yield optimization while limiting the waste of water and other resources.

Hamamatsu Photonics' grating-based portfolio includes instruments for different integration and analysis requirements. One example is a fingertip-sized spectrometer whose longest dimension is 20 mm, integrating a silicon sensor sensitive from 190 nm up to 1,000 nm depending on the model. Its compact form factor is enabled by a proprietary CMOS process that etches the input slit directly onto the linear image sensor. This analog device can be free-space based or fitted with an SMA connector for fiber coupling, giving integrators flexibility in optical and mechanical design. At the module level is a plug-and-play mini-spectrometer module offering USB or Ethernet connectivity, fiber input, and higher sensitivity and resolution. It is available in cooled and uncooled versions across ranges from UV to SWIR.

Solving the Value Problem with Shortwave Infrared

The importance of composition over appearance becomes clear in two scenarios. Consider two neighboring farmers growing the same crop, one harvesting early and one late. If both are paid by weight, the farmer selling drier, later-season produce is valued less despite delivering greater nutritional value due to a lower water content and therefore lower weight than the early-harvest produce. As the weight reflects water content as well as dry matter, sugar, starch, or fiber, the other farmer will be paid more, but not necessarily because of a higher-value crop. Similarly, an animal farmer wants full control over the nutritional values of feed and assurance that its advertised characteristics hold true. Both problems point toward the same tool: a SWIR mini spectrometer.

Sharing the compact form of the aforementioned silicon device, it uses an indium gallium arsenide (InGaAs) detector sensitive from 850 nm to 1,700 nm—the fingerprint region where absorption features of sugar, protein, starch, and water appear. At 20 mm with a spectral resolution near 15 nm, it integrates temperature measurement for reliable readings as conditions change and offers free-space or fiber configurations with an electronic shutter. Extending analysis from a single point to many pixels enables imaging, widely used to sort foreign objects from food. Stones that are difficult to distinguish among coffee beans by eye appear as dark spots under SWIR and can be identified thousands of times per second as product falls or moves along a conveyor. The same imaging can reveal bruises in fruit invisible under normal light, supporting quality-based sorting.

Fourier Transform Spectroscopy

Fourier transform spectroscopy (FTS) routes reflects light through a beam splitter into two arms: one reflecting off a fixed mirror, the other off a movable mirror. Recombined and directed to a SWIR InGaAs detector, the beams produce an interference pattern that varies with mirror position, forming an interferogram. A fast Fourier transform then yields the absorption spectrum, in this case spanning 1,000 nm to 2,500 nm. This technique supports fingerprinting of food origin, composition, and processing history.

These are useful for fraud prevention with high-value goods such as honey, olive oil, and wine. It also supports soil analysis and food quality assessment. A dairy processor, for example, might blend milk from multiple suppliers to maintain consistent final characteristics.

Flow Cytometry

Hamamatsu's FTS instrument covers 1,100 nm to 2,500 nm, and, in its latest iteration, reaches 275 frames per second. This speed is fast enough to classify sugar from artificial sweetener, in real time, on a conveyor belt. An advantage of the approach is it avoids the issues between resolution and signal-to-noise ratio inherent to grating instruments, where narrowing the entrance slit sharpens resolution but sacrifices light.

A demonstration on almonds, cashews, and walnuts illustrated the workflow. After acquiring full absorbance spectra, a discriminating spectral window was isolated, a second derivative applied, and principal component analysis used to classify the samples, which cleanly separated. Repeating the test on powdered nuts produced even clearer separation, as grinding yields a more homogeneous sample.

The Fabry-Pérot Interferometer (MEMS-FPI)

A third instrument, the Fabry-Pérot interferometer, uses two mirrors. One is fixed at the bottom and one adjustable by applied voltage to form a variable filter that sweeps the air gap between them before light reaches an InGaAs detector. This principle was applied to distinguish wheat, rice, and buckwheat flours. The resulting handheld analyzer integrates its own light source, operates across three ranges of roughly 300 nm each covering 1,350 nm to 2,150 nm, and is USB powered and controlled. Its affordability suits applications such as moisture detection near 1,450 nm, sugar content, and fruit ripeness measured by Brix value, with potential in process control such as acetic fermentation monitoring. When run on the same unprocessed nuts, it again produced a valid classification despite its narrower range.

Gas Sensing in the Mid-Infrared

Beyond 3,000 nm, in the mid-infrared, spectroscopy becomes useful for gas measurements which is relevant to farming for three reasons. First, new EU regulation is expected to set maximum ammonia and methane levels in animal barns for livestock safety. Second, methane emissions from livestock, which are a significant greenhouse gas source, face monitoring requirements, and carbon trading schemes may prompt carbon dioxide (CO2) measurement.¹ Lastly, with the rising cost of gas and some countries not having access to natural gases, they are turning to biogas production. Biogas produced from agricultural by-products and manure benefits from quantification.

Non-dispersive IR spectroscopy handles these measurements by passing light through a gas cell to a detector fitted with a bandpass filter, which is matched to a target gas's absorption band. When there is no gas in the chamber, a baseline signalis indicated. When CO2 enters the gas chamber, it absorbs part of the light at 4.3 μm, decreasing the detector signal. With this decrease, not only is gas identified, but it is also quantified based on the concentration of gas in the environment.

Hamamatsu offers mid-infrared detectors based on indium arsenide antimonide and optimized LEDs for gases including carbon dioxide and methane. Compared with electrochemical methods, the photonic approach offers higher sensitivity and selectivity, fast responses, and minimal maintenance; it can run for years after a single calibration, making it well suited to remote or battery-powered deployment.

Conclusion

Within agriculture, the properties that determine a product's value or risk are largely invisible to the eye. None of these properties can be judged reliably by weight, color, or appearance, yet those visible proxies have long governed how crops are priced and how quality is assessed. Grating-based spectrometers, Fourier transform instruments, the Fabry-Pérot analyzer, SWIR imaging, and mid-infrared gas sensing each convert one of these hidden properties into a measurable value. These are at a scale and cost increasingly suited to drones, conveyor lines, and handheld field devices. With science already in place, the remaining challenge is translation: matching a grower's specific problem to the appropriate photonic technique and bringing it to market at scale. To help bridge this gap, Hamamatsu has established a European Pilot Line team that supports customers from feasibility studies and specification definition through optoelectronic module design, prototyping, and industrialization. By combining Hamamatsu's expertise in light emitters and detectors with the application knowledge of customers, the Pilot Line service helps accelerate the path from product concept to mass manufacturing while reducing development risk, time, and cost.

Want to Learn More? Watch the Full Webinar.

Dive deeper into the topic and hear directly from our expert. The full webinar recording is now available to watch on demand whenever it suits you.