Boosting net profit to 27.8%, Koidra won the Autonomous Greenhouse Challenge, again

May 10, 2023

Executive Summary Outplaying other international competitors, a joint team led by Koidra won the 3rd…

Executive Summary

Outplaying other international competitors, a joint team led by Koidra won the 3rd Autonomous Greenhouse Challenge (AGC) in 2022 by optimizing the final net profit to 27.8%. Koidra’s technology was also at the core of team Sonoma’s victory in the previous AGC challenge after generating the impressive net profit of €5.93 per one square meter per crop cycle.

Customer Overview

The Autonomous Greenhouse Challenge is an international competition organized by the Netherlands-based Wageningen University & Research (WUR)–the world-leading institution for Agri-Food Tech. In its third year, the challenge witnessed the participation of 46 teams from 24 countries. Team Koala consisting of agtech startup Koidra and researchers from Cornell University and Rutgers University emerged as the winner. 

Challenges

The goal of the competition was to optimize net profits by producing a maximum number of good quality heads of lettuce with the lowest feasible input of resources such as energy and CO2. The teams had to grow lettuce with a target plant weight of 250 grams, and the quality was judged based on factors such as leaf tip burn and other deformations.

They also had to work with a high-tech greenhouse compartment at Wageningen University & Research in Bleiswijk. The resource use (e.g. heating energy, electricity, CO2) was measured during the growing period, and operational costs were calculated.

Koidra’s Solution

The Koala team lead by Koidra developed computer vision and machine learning algorithms to detect growth parameters including diameter, height, weight, and leaf area, and to control the climate in a virtual greenhouse by automatically determining set points for ventilation, heating, lighting, etc. in order to grow a virtual lettuce crop and maximize net profit.

Results & Benefits

Koala team was named the winner of the competition after building an algorithm that recognized high-accuracy lettuce images and estimated the accurate growth parameters.

The team achieved 400g per head of lettuce, which greatly surpassed the target of 250g per head (Figure 1) with the least use of resources (Table 1).

The Koala team was the only AI team to outperform the reference Dutch Grower by 27.8% in net profit (Table 2).

Figure 1: Cummulative total production of lettuce per team per crop cycle. 
Table 1: Resource use per kg sellable lettuces per team. Heat, electricity and CO2 use differed largely per team.
Table 2: Net profit and total cost calculation from income sellable lettuces minus fixed costs and operational costs (heating, electricity, CO2, intervention to the algorithm) per crop cycle.

Koidra’s participation under the umbrella of team Koala in the AGC and its success in delivering autonomous growing technology have been a significant achievement for the company. The Autopilot solution has demonstrated its capability to provide quality crops, boost energy usage efficiency and reduce operating expenses. Koidra’s innovations have provided a competitive advantage that can benefit commercial lettuce production and the agricultural industry as a whole.

About Koidra

Our interdisciplinary approach combines machine learning, IoT, Math, Physics, and expert knowledge to enhance operational precision and efficiency beyond what humans can do alone. Our intelligent automation solutions enable growers and operators to increase their span of control and reduce dependence on human expertise. By doing so, we make industrial automation smarter, easier to deploy, easier to evolve, and hence more scalable.

Sources