Around 330 million male chicks are killed in the EU every year, because they lay no eggs and put on too little meat to pay for their feed. Germany and France have already banned the practice, and other Member States are pushing for an EU-wide ban, which leaves hatcheries with a question they have to answer before anything hatches. A German team funded by the European Innovation Council shines light through the shell and lets an algorithm read the answer - a mix of ethics, optics and code that is worth asking yourself about before you pick a degree.
An estimated 330 million male chicks are killed in the EU every year. They hatch, they are sorted, and they are gone within hours, because the industry has no use for them. The figure is large enough that several Member States have called for an EU-wide ban on the procedure, and two of them did not wait for it: culling is already prohibited in Germany and France.
A ban, however, removes the option of sorting chicks after they hatch. The sorting has to move earlier, inside the egg, and that is a measurement problem rather than a legal one.
The economics behind the number are blunt. Laying hens are bred for eggs, not for muscle, so their brothers neither lay eggs nor produce enough meat to be profitable. Feeding and housing them costs money that never comes back, which is why the sector treats them as waste rather than as a product. Next to the ethical objection there is a second one that is easier to put on a spreadsheet: this is a significant waste of biological material at industrial scale.
The EIC-funded Omegga project attacks the problem with spectroscopy and machine learning. The methods used for sexing chicks today are invasive, inefficient and costly, and several of them involve puncturing the shell. Omegga's system does not touch the shell at all.
"The eggs are screened using light and the emitted spectrum measured," says Omegga CEO Katharina Hesseler. "These measurements are repeated over several days and analysed using AI to identify subtle sex-specific changes in the egg, including gender-specific developments in the embryo."
Timing is what decides whether the method counts as an ethical answer or merely a technical one. The system determines the sex of the embryo while it is still inside the egg by day six of incubation, well before the development of any pain perception. Male and non-viable eggs can then leave the incubator early and go somewhere useful: the project names vaccine production and animal feed. Nothing has to be culled at all.
Reading a reliable signal through an intact shell was the hardest scientific part of the work, and hardest of all during the first days of embryo development, when the differences are smallest. Eggs are not standardised objects either. They vary in size, shape, colour and shell thickness, and every one of those variables muddies the measurement.
"We overcame this with our time-series approach," explains Hesseler. "By comparing each egg against its own earlier measurements, we can reduce biological noise and amplify the sex-specific signal."
Each egg, in other words, becomes its own reference point. It is a habit worth borrowing in any field where the objects being measured refuse to be identical.
The technology is designed as a plug-and-play module that can be retrofitted into standard industrial incubators, so hatcheries of any size can adopt it without rebuilding infrastructure they already own. That detail decides adoption more often than accuracy does: a method that demands a new building tends to stay in the laboratory.
"Nothing is cheaper than light, which is affordable, fast and precise," Hesseler says. "Furthermore, it's non-invasive and purely optical, meaning no chemicals or needles are required and no additional personnel."
The target the company has set itself is correspondingly large. "We aim to eliminate the culling of up to 100 million male chicks annually," Hesseler points out. "This aligns with the EU's strategic goals to improve animal welfare, promote sustainable agricultural innovation and support a circular economy in food production. Our ultimate objective is to make culling-free egg production the global standard."
An optical sensor plus a trained model is a generic combination, and the project says so openly. Hesseler points to three directions in which the same core platform could be used:
"By building on the same core platform, we can expand into new markets over time, creating a scalable business with opportunities for significant long-term growth potential," Hesseler concludes.
Omegga uses light and machine learning to read the sex of a chicken embryo through an intact shell by the sixth day of incubation, early enough for male eggs to leave the incubator and be repurposed for vaccine production or feed. The method is non-invasive, it retrofits into hatcheries that already exist, and it is aimed at a practice that currently ends around 330 million lives a year in the EU alone. The science behind it is optics, biology and data analysis in one device, and none of the three works without the other two.
5 questions · one minute · nothing is saved
EN
CS
DE
ES
FR
HU
IT
PL
PT
РУ
SK
TR
УК
AR
中文
