Akkodis Data Analytics Solutions Help Client Achieve AgTech Breakthroughs
The client is bringing the era of computational agriculture to life - the invisible links supporting life on Earth. An agriculture system optimized for productivity and simplicity comes with risks, as agriculture is unpredictable.
The project is building on breakthroughs in artificial intelligence, machine learning, simulation, perception, the IoT, large unmanned rovers, and robotics. This project aims to build newer software, hardware, and a data-driven platform that bring together diverse sources of insight to manage the staggering complexity of farming.
As part of the client's innovation initiative, the Akkodis team needed a new approach to working with fast-changing, unpredictable technologies. This approach requires a 10x thinking mindset, rapid iteration, commitment bias, failure aversion, and quick experiments. Data management is critical when there are condition variations in the environment.
Building a complete picture of the field, the Akkodis team supported enhancing and optimizing operations through various prototypes and Mineral rovers, gathering high-quality images of each plant. Then the data is combined with other imagery datasets (e.g., satellite images), identifying patterns and analyzing the insights into how plants interact with the environment.
The team has been developing and implementing these newer technologies that offer growers insight into what’s happening in their fields, right down to the individual plant level. We have evolved the camera and sensing technology, mechanical and electronic design on buggy prototypes so that the program can now do tasks that are otherwise impossible for humans, like counting the individual buds on every raspberry cane or accurately estimating the number of soybeans in a field.
As our prototypes evolved, the team had also come up with a simplified, easy-to-use mini-plant buggy that could be shipped in a small box and used anywhere: something that was never expected to evolve.
Extending this further, the team supports experimentation for Berry QA using newer technologies, certifying quality standards such as freshness, ripeness, and appearance to meet the USDA quality standards.
The technology development in the Berry QA initiatives involves:
- Harvesting data collection (using an innovative mobile sensors-based device) supporting computer visions subset of AI image acquisition
- Data preprocessing
- Sorting, assessment, and data analytics to check the quality and provide trusted standards certification labeling.
Industry: Manufacturing & Logistics
Technologies & Certifications used:
Google Cloud Platform