Report
Teaser, summary, work performed and final results
Periodic Reporting for period 1 - ROMI (RObotics for MIcrofarms)
Teaser
All over Europe, young farmers are starting small market farms and direct sales businesses. These farms can be found both in rural, peri-urban and urban areas. They grow a large variety of crops (up to 100 different varieties of vegetables per year) on small surfaces (0.01 to...
Summary
All over Europe, young farmers are starting small market farms and direct sales businesses. These farms can be found both in rural, peri-urban and urban areas. They grow a large variety of crops (up to 100 different varieties of vegetables per year) on small surfaces (0.01 to 5 ha) using organic farming practices. These farms have proven to be highly productive, sustainable and economically viable. However, a lot of work is done manually, resulting in physically challenging work conditions.
ROMI will develop an open and lightweight robotics platform for these microfarms. We will assist these farms in weed reduction and crop monitoring. This will reduce manual labour and increase the productivity through advanced planning tools. Thanks to ROMI’s weeding robot, farmers will save 25% of their time. This land robot will also acquire detailed information on sample plants and will be coupled with an aerial robot that acquires more global information at crop level. Together, they will produce an integrated, multi-scale picture of the crop development that will help the farmer monitor the crops to increase efficient harvesting. For this, ROMI will have to adapt and extend state-of-the-art land-based and airborne monitoring tools to handle small fields with complex layouts and mixed crops.
To achieve this, we will: (i) develop and bring to the market an affordable, multi-purpose, land-based robot, (ii) develop a weeding app for this robot that is adapted for organic microfarms, (iii) apply advanced 3D plant analysis and modelling techniques to in-field data acquisition, (iv) integrate these analysis techniques in the robot for detailed plant monitoring, (iv) integrate these techniques also in an aerial robot for multi-scale crop monitoring, (v) extend the robot with novel, adaptive learning techniques to improve sensorimotor control of the plant monitoring app, and (vii) test the effectiveness of our solution in real-world field conditions.
Work performed
ROMI choose to set mechanical weeding as a short-term priority because it is a time-consuming and physical task in organic farming. Indeed, a weeding application is often the first item on the wish list of organic farmers. Moreover, such a tool could convince more conventional farmers to transition to organic practices, providing an appealing and credible solution to accompany the necessary cut-off of herbicides. ROMI’s second short-term application is a static, indoor 3D plant scanner. The scanner is a stepping stone to develop the outdoor phenotyping application. We believe that the scanner is a useful object in itself that can be proposed to biology labs interested in precise and quantitative plant architecture phenotyping.
Our long-term vision is an intelligent robotics platform for farming, that collects data from the field, informs farmers about development and health status of crops, watches weed growth and directly operates on culture beds (e.g. mechanical weeding). This involves appropriate sensors and dedicated algorithms that give the robot the advanced capability of observing plants, plant populations and the whole field. We choose to observe the shape of the plant (phenotype) over time using techniques from computer vision, 3D vision, plant modeling, active vision, and AI. This requires a novel combination of adaptive 3D imaging techniques and plant models that is robust and fast enough to function in outdoor conditions. The platform will build up an accurate and informative picture of what happens in the field by combining a top-down with a bottom-up approach: an airborne device snapshots a large-scale view of the field while a ground-based rover collects detailed information from individual plants.
Final results
The ROMI project makes a big deal of sustainability for Tomorrow’s agriculture. Organic microfarms are promising sustainable agricultural systems, but they still need to convince a majority of their ability to feed the world while providing good incomes to farmers. We believe that new and adapted technologies can help these systems to solve the conundrum of combining productivity, reduced environmental impacts and economical viability. Although agroecology claims that increasing the level of biodiversity and polyculture can increase productivity and resource-use efficiency with nutrition and environmental benefits, efficient management of such complexity may hinder the adoption of such practice. For instance, most of the modern agricultural mechanics has achieved tremendous yield increase through homogenization of agrosystems, allowing efficient automation of basic tasks.ROMI makes an opposite bet: matching the needs of more complex farming environment with a more complex computational environment. Our aim is to develop flexible, smarter, adaptive tools to manage complex and biodiverse environments. The increased complexity can be compensated for by using tools with advanced sensing and modeling capabilities, by increased collaboration between farmers and scientists and by creating farming communities gathered around innovations for a numeric agriculture. We hope that during the ROMI project, we will be able to develop a proof of concept of this approach.
Another important feature of ROMI is that all the tools (software and hardware) will be made openly available under a free license. We believe that this is the best approach to have a wide as possible impact. It facilitates the access to our results for the many small farms worldwide and facilitates the collaboration between farmers, scientists, engineers, and industry. A start-up will be created to bring the open-source robotics platform to the market.
Website & more info
More info: https://romi-project.eu/.