SWIR LED Food Sorting Application

The problem as it presented itself

The customer arrived with a working prototype and a production problem: the LEDs that passed the bench test were not the ones that survived the line. Nothing in the data was out of spec — the failures were spread across batches rather than concentrated in one lot, which ruled out a supplier problem and pointed at the process instead.

Three measurements separated the causes. First, forward voltage at a fixed low current across a sample of each batch, which showed whether the die set was consistent. Second, case temperature under the real drive condition, which showed whether the thermal path matched the assumption in the design. Third, the reflow profile logged on the line, which showed whether the assembly process stayed inside the qualified window.

The first two came back clean. The profile log showed the peak temperature sitting close enough to the limit that normal line-to-line variation pushed parts past it. This is the most common way a qualified component becomes an unqualified assembly, and it is invisible on the datasheet.

• Measure the process, not just the part. A component that meets its specification can still be assembled outside its qualification.
• Log the profile, do not assume it. Reflow profiles drift as boards and fixtures change.
• Look at failure distribution first. Failures spread across batches point at the process; failures concentrated in one batch point at the supply.

Once the profile was corrected and re-qualified, the field return rate dropped to the level the component datasheet predicts. The component never changed — only the process around it.

Nut and grain processors face a stubborn inspection problem: shell fragments, insect damage, and foreign matter that look identical to good product under visible light. This project shows how 1470nm SWIR LEDs solved it — raising foreign-matter rejection accuracy to 99.2% on a high-throughput sorting line.

Application Background

A nut processing company supplying retail brands in Europe and North America needed to upgrade its optical sorting line. The existing color-camera system reliably removed visibly discolored kernels, but three defect classes kept escaping to the customer’s incoming inspection:

• Shell fragments — thin, pale pieces that match almond flesh under RGB imaging
• Concealed insect damage — internal cavities with no external color signature
• Plastic and rubber debris — from conveyor wear, often similar in color to product

Each escaped defect risked chargebacks and audit findings under the customer’s food-safety program. Manual re-inspection of flagged lots was costing the plant 40+ labor hours per week.

Engineering Challenge

The physics of the problem pointed to the short-wave infrared band. Water and organic materials interact with SWIR light very differently than with visible light:

• At 1470nm — close to a strong water absorption peak — the moisture contrast between sound nut flesh and damaged or foreign material produces a dark-bright signature invisible to RGB cameras
• Shell, plastic, and insect-damaged tissue show clearly differentiated reflectance against healthy kernels at this wavelength

Three constraints shaped the illumination design: the sensors needed uniform line illumination across a 1200mm scan width; the LEDs had to deliver sustained radiant power under continuous duty at 40°C ambient inside the sorter housing; and the source lifetime had to match the machine’s 5-year service interval to avoid mid-season teardowns.

Solution & Key Components

Queendom supplied SWIR LED arrays built on the 3535 high-radiance platform, configured for line-scan illumination:

Integration notes for sorting-machine builders: the modules mount behind protective sapphire windows (SWIR transmits well; glass absorbs partially), and the constant-current zones map one-to-one to the camera’s pixel banks so intensity can be trimmed per zone during commissioning.

Results

After a two-week validation run, the SWIR channel caught shell fragments and plastic debris that the RGB stage missed, while the false-reject rate stayed below the 0.5% contract threshold — keeping yield losses acceptable.

Recommended Products

3535 Infrared LEDs

High-radiance ceramic platform for SWIR array modules

LED Diodes Product Line

Full infrared, visible, and UV component portfolio

AOI Machine Vision Case

Related industrial inspection application

Frequently Asked Questions

Why 1470nm instead of 1550nm for food sorting?

Both wavelengths sit in the SWIR water-contrast band. 1470nm offers a practical balance: strong moisture contrast, available high-radiance LED sources, and better sensor compatibility than longer SWIR wavelengths, which increasingly require costlier InGaAs detector options.

Can SWIR LEDs replace RGB cameras entirely?

No — the two channels are complementary. RGB removes visibly discolored product; SWIR catches moisture, shell, and foreign-material contrast invisible to color imaging. Most modern sorters run both channels in sequence.

How are the arrays protected in wash-down environments?

Modules sit behind sapphire or SWIR-transmissive windows in IP65-rated housings, with gasketed cable glands — the same enclosure discipline used in the food-washdown lighting described in our waterproof lighting applications.