High-tech agricultural tech company is at the forefront of AI usage in helping swine businesses work smarter and more efficiently.

by Andrew Joseph, Editor

Conception Ro-Main Inc. is a high-tech ag company headquartered in Saint-Lambert-de-Lauzon, Québec, Canada. Established in 1999 by brothers Robert and Germain Labrecque, it’s the amalgamation of their names “Ro-” and “-main,’ which smartly denote the company moniker.

The company’s first technological invention was the development of the Contact-O-Max, a remote-controlled boar cart that revolutionized swine reproduction by facilitating large-scale artificial insemination.

Over the years, Ro-Main has evolved into a global leader in precision livestock farming, particularly in the swine industry. The company specializes in creating intelligent, practical technologies that enhance farm productivity and animal and worker welfare. The company is housed in a 50,000-square-foot facility, with 48 employees.

In February 2025, Ro-Main joined the IMV Technologies Group, a global leader in assisted reproduction biotechnologies, for a partnership expected to accelerate Ro-Main’s growth and further its mission of providing sustainable, high-tech solutions for animal farming.

Ro-Main continues to operate with a multidisciplinary team of veterinarians, agronomists, data scientists, engineers, and computer vision experts, all dedicated to supporting farmers worldwide in achieving more efficient and ecological protein production through high-tech solutions.

Ro-Main's SmaRt Counting - its latest AI technology tool - with the integrated SmaRt Weight technology. Photo by Conception Ro-Main Inc.

Its current product portfolio includes:

• SmaRt Counting: an AI-powered system that automatically counts pigs with high accuracy, aiding in inventory management.
• SmaRt Breeding: a tool that monitors sow behaviour to determine optimal insemination times, improving reproductive efficiency.
• Contact-O-Max: a self-propelled, remote-controlled boar cart that optimizes sow sexual stimulation by ensuring sustained and precise snout-to-snout contact, improving reproductive performance.
• Hercules’ Arm: a self-propelled carcass cart designed to assist in the safe and ergonomic removal of deceased animals, especially large sows, from tight spaces, enhancing worker safety.
• Companion Cart: a self-propelled carcass cart designed to assist in the safe and ergonomic removal of deceased animals, especially gilts and finisher pigs, enhancing worker safety.
• No Backdraft: a ventilation solution that prevents air backflow, maintaining stable and microbe-free barn environments.

Benchmark magazine talked with Jacquelin Labrecque, Ro-Main’s Chief Executive Officer, who joined the company in 2011. Born and raised on his family’s pig farm, Labrecque pairs firsthand knowledge of swine farming with his academic background in mechanical engineering and artificial intelligence. After rising through the ranks in both technical and leadership roles, he became CEO in early 2024.

“In 2008, we stepped into the realm of data-driven decision-making, integrating software and algorithmic technologies into our family’s pig farms,” stated Labrecque. “It started with PigWatch, a motion sensor-based Italian system designed to predict the optimal time for insemination, a product technology that Ro-Main eventually acquired.”

Labrecque explained to the magazine that he discovered a passion for artificial intelligence (AI) while pursuing his master’s degree. “Combining AI with our work on PigWatch revealed just how powerful these technologies could be in transforming swine farming by enhancing human decision-making and automating key processes.”

He explained that while the company has always been driven by innovation, 2014-2015 marked a turning point. It was during this time that Ro-Main reinvented itself, embracing modern artificial intelligence - a set of tools previously untapped in livestock farming.

This strategic shift placed them at the forefront of technological advancement in the swine industry, pioneering new ways to optimize production and animal care.

“The evolution of our flagship products - PigWatch, SmaRt Breeding, and SmaRt Counting - reflects not just technological progress, but a deep understanding of how to bring AI into one of the most complex and traditional industries: livestock farming,” he noted.

While the company’s PigWatch technology was groundbreaking at the time, Labrecque explained that as deep learning and computer vision evolved, there was an opportunity to transform the foundation of the system.

This led to the birth of SmaRt Breeding, which Labrecque called a reimagined, camera-based system that gathers far richer data using a single camera to monitor four animals simultaneously.

It reduced hardware costs and unlocked much higher precision.

“From there, we focused on leveraging farm-collected data to improve the AI’s accuracy, especially for single-dose insemination,” he explained. “It was a major milestone that truly elevated SmaRt Breeding into a predictive, continuous learning platform.”

The SmaRt Counting product followed a similar path but tackled a different challenge: accurate pig counting. The first hurdle was to ensure the AI could perform with high precision in vastly different farm environments, such as different lighting conditions, pig sizes, breeds, and behaviours. “It took two years of refining to get it right, and that was our first major milestone: a working, reliable counting algorithm.”

Next came packaging. After starting with a consumer-grade computer in a robust case, Ro-Main built its own Edge Intelligence Platform (EIP), an industrial-grade supercomputer in a rugged enclosure, designed entirely in-house to be farm-ready. “This iteration,” said Labrecque, “allowed us to shift the focus from technology to value delivery.”

But, as he pointed out, even with powerful AI and hardware, adoption only happens if it fits into existing SOPs (standard operating procedures) and workflows. “We developed new features like usage compliance tools that act as auto-training systems for farmworkers while also alerting managers about potential usage issues or even animal welfare concerns - like pigs piling up during counting.

“The goal was simple: make the tool seamless, intuitive, and valuable for every stakeholder, from the person counting the pigs to the business owner analyzing performance,” commented Labrecque.

“But, through this journey, one truth became clear. Building AI for this industry is about much more than AI,” he explained. “It’s about being a pioneer, yes, but that means guiding the change, not just sparking it. It’s our responsibility to help farmers adopt the technology, not just use it.”

Labrecque said that Ro-Main is continuing that mission with its next product, SmaRt Weight, which he said would debut at the 2025 World Pork Expo, June 4-5, 2025, in Des Moines, Iowa.

“The SmaRt Weight is an add-on to SmaRt Counting, which provides not only pig counts but also average weights and coefficient of variation - with zero extra effort,” he pointed out. “This gives farmers instant insight into the uniformity of a group, which in turn reflects how well the animals were managed during early life.

“That data,” he continued, “helps save time and money and can guide decisions about treatment, feeding, and forecasting through to market. It’s not just about better data; it’s about actionable intelligence, available instantly, at scale.”

He said that AI, and particularly deep learning, has opened a new era in swine monitoring “by allowing us to capture and analyze unstructured data - like images, video, or even sound - 24/7, just as if we had human senses present in the barn at all times.”

Cameras and microphones can now continuously observe animal behaviour, while AI interprets the data in real-time to detect patterns, count pigs, and monitor their well-being.

“One of AI’s greatest strengths is its ability to learn from variation,” revealed Labrecque. “In a pig barn, for example, where lighting conditions, animal sizes, and movement can vary widely, deep learning allows our systems to adapt and perform with a high level of precision.

That said, AI is not magic, nor is it a human brain. Because it’s a powerful tool, it needs the right conditions and training to perform reliably. “That’s why, when we design our products, we don’t just rely on AI. We take a holistic approach,” he noted.

“We identify edge cases—unusual but important situations—and we adapt our hardware and processes to reduce the burden on the algorithm. Sometimes that means improving lighting, adjusting camera angles, or engineering mechanical features that create more consistency in the environment,” offered Labrecque. “Ultimately, it’s the combination of human ingenuity and AI capabilities that makes these systems work so well.

“The goal isn’t just to make smart tools,” he summed up, “it’s to make smart tools that work on real farms, under real-world conditions.”

For more information on Ro-Main, visit www.ro-main.com.

---

---