Suzanne Saliba Baron on Connecting Data and Smart Implements for Autonomous Farming
From vineyard trials to real-time machine learning, AgreenCulture is making autonomy in agriculture a practical reality. In this interview, we speak with Suzanne Saliba Baron, Marketing and Communication Director at AgreenCulture, one of the partners behind the Robs4Crops project.
Suzanne shares insights on how her team integrates smart implements, robotics, and data to make precision agriculture more resilient, scalable, and farmer-friendly.
Exploring Our Guest’s Motivation and Background
What was your initial motivation to involve yourself in the AgTech world, and what motivated you to stay and grow in this field?
Our journey into AgTech began with a clear observation: many farmers across Europe face increasing pressure to manage their farms alone, often with limited resources. To remain resilient, they need access to smarter, more autonomous tools. That’s what initially drove us to develop innovative solutions, including an AGC autonomy kit for an autonomous tractor and UAV, which integrates the guidance and positioning unit, a vision system (comprising cameras), and LED indicators for real-time status feedback.
At the beginning of AgreenCulture, we were working in space technology, which gave us a strong understanding of high-precision positioning systems. We quickly realised how critical these technologies could be in agriculture: a sector where every centimetre, every hour, every decision counts.
We also saw strong interest from our partners, including Kubota, Pellenc, and Khun, all of whom confirmed both the relevance and urgency of our work. The growing need for interoperable, technology-agnostic tools in agriculture reinforced our decision to stay in the field and continue innovating.
Developing Autonomous and Connected Systems for Safer, Smarter Farming
You’ve been deeply involved in developing autonomous systems for agriculture at AgreenCulture. How do you see the role of structured, accessible agrifood data in supporting the safe and efficient deployment of autonomous machinery in the field?
Developing autonomous machines is just one part of the challenge; ensuring their safety is absolutely essential. That’s why at AgreenCulture, we’ve made it a priority to design certified geofencing systems called Safencing, which guarantee both the security and resilience of autonomous operations in the field.
But safety alone isn’t enough. For farmers to adopt these technologies, efficiency and ease of use are just as critical. That’s where PAM comes in, our Platform for Autonomous Management, designed based on farmers’ real experiences and practical needs. It enables farmers to remotely monitor and manage their autonomous machines through a user-friendly cloud-based interface.
One of PAM’s greatest strengths is its adaptability: it integrates seamlessly with a wide range of existing industrial platforms from our partners, making it a versatile tool for modern, data-driven agriculture.
One of STELAR’s goals is to make it easier to link datasets with data processing workflows for Machine Learning and AI applications. From your perspective, what types of datasets are most crucial for smart implementations in precision agriculture? Are there gaps or bottlenecks you regularly encounter when sourcing or using data?
Farmers remain the true experts in precision agriculture – they know best how to assess the quality of work done in the field and how to adapt operations in real-time. That’s why, in our view, the most valuable datasets are those that capture and reflect the farmer’s decision-making process, whether it’s through task execution data, environmental conditions, or machinery behaviour.
AI applications shouldn’t aim to replace the farmer, but rather to support and enhance their work. That means developing systems that can learn from their expertise, replicate best practices, and adapt to real-world variability in the field.
One of the key issues we face is access to high-quality, structured, and interoperable datasets. Much of the data collected today is fragmented or not standardised across equipment and platforms, which makes it harder to build scalable, efficient AI tools. Solving this requires not only better data infrastructure but also stronger collaboration between farmers, developers, and manufacturers.
Interoperability and the Value of Connected Systems
AgreenCulture works at the intersection of robotics, AI, and on-the-ground farming realities. Can you share a concrete example where improved data access or integration significantly impacted the performance or reliability of an autonomous farming solution?
A great example comes from our work in vineyards, where the field reality and crop management are highly specific and require tailored solutions. In one of our use cases, we deployed an autonomous system that uses AI not just for navigation, but also for monitoring the quality of work done by the attached implement, for example, weeding.
Thanks to improved access to field data and better integration of sensor feedback, the system was able to evaluate the work autonomously, reducing the need for the farmer to follow the machine and double-check everything. This significantly lightened their workload while ensuring consistent quality of operations.
This was particularly evident in the Robs4Crops project, where combined robotic platforms, AI-driven decision tools, and data from real-world farming scenarios are combined to deliver a more intelligent and reliable autonomous solution.
In many instances, we’ve seen that combining data from different sources is key to real innovation. How important is data interoperability and standardisation for your work, and what advice would you give to projects like STELAR working to improve this?
Data interoperability and standardisation are central to our approach at AgreenCulture. With PAM, our cloud-based platform for fleet management and autonomous machine monitoring, one of our key objectives is to ensure seamless integration with other technologies already used by our partners.
Rather than asking them to redevelop systems from scratch, we focus on making PAM easy to plug into existing infrastructures, thanks to a standardised communication interface. This allows our partners from OEMs to digital agriculture platforms to enhance their offer with autonomous capabilities without disrupting their core architecture.
Access to multiple, diverse datasets is also a major asset. It helps us develop more robust algorithms, better suited to varied terrains, crop types, and operational contexts. That’s why we strongly support efforts like STELAR that aim to facilitate data sharing and harmonisation across the agricultural ecosystem. Our advice? Design for modularity, listen closely to end users, and ensure that standards are co-developed with both the tech industry and the farming community.
Smart Implements and the Road Ahead for AgreenCulture
STELAR also explores tools for extracting information from unstructured sources, like food safety reports. From your view, could such structured outputs be useful to agri-tech companies like AgreenCulture? What kind of intelligence would you find most valuable to integrate into smart systems?
At AgreenCulture, we design our technologies with a “farm to fork” approach in mind. That’s why we see real potential in using structured outputs from unstructured sources – they could help us align our solutions more closely with the entire agricultural value chain.
Such information could support the development of smarter, more connected systems that not only operate efficiently in the field but also respond to broader quality and traceability requirements across the food system.
Looking ahead, what do you think are the most exciting opportunities for connecting agricultural data and smart implements?
Smart implements represent the future of AgTech, and the real value lies in how we connect them through data to deliver actionable, intelligent, and scalable solutions.
At AgreenCulture, we’re already seeing this in action through projects like Smart Droplets, where we combine real-time data with precision application systems to optimise resource use and minimise environmental impact. Another great example is AGRARSENSE, which explores new ways to fuse sensor data, machine feedback, and AI for enhanced in-field decision-making.
Today, we use data primarily to ensure safety, monitor vineyard conditions, and support the autonomy of our machines. But tomorrow, we see opportunities for deep integration: using live data to adapt and implement behaviour in real time and anticipate agronomic needs. The real opportunity lies in making all these systems communicate and learn together, unlocking a new era of responsive, interconnected, and farmer-centric agriculture.
Conclusion
Suzanne Saliba Baron offers a compelling perspective on how AgreenCulture bridges robotics, AI, and smart implements to deliver practical, modular solutions in agriculture. Her insights reflect a commitment to both farmer needs and long-term sustainability in the agri-tech ecosystem.
To read more interviews and updates from STELAR, visit our Blog or follow us on LinkedIn.
