Singapore-based Entoverse, a startup delivering cutting-edge AI-driven digital farm management solutions for the fast-growing insect farming sector, announced today its trials to integrate Chat Generative Pre-Trained Transformer (GPT) intelligence into its Entograte ERP system.
Common challenges in insect farming include mortality and disease arising from environmental conditions or insect feed, which typically cause 25-50% losses, as well as human error when managing populations of millions of insects. With an ERP and access to 2 supercomputers, the Entograte system provides powerful AI-driven solutions to address key production challenges in insect farms, increasing efficiency in insect egg production and larvae rearing.
CEO of Entoverse, Rudolf Davidov, said: “Our species-specific ERP system will make day-to-day operations of insect farms more efficient and automated, while our AI-driven solutions will help resolve the challenge of early mortality and optimising product quality.”
Following the introduction of GPT intelligence towards the end of 2022, Entoverse’s latest innovation seeks to apply the technology to overcome the challenge of human management of increasingly large and complex insect farms. Fast communication and reaction to changes in farming conditions are key, and FarmGPT is being integrated into Entograte’s chatbot function to achieve this.
It will deliver data-driven actionable insights that address insect cultivation challenges in real-time. FarmGPT will be trained by a combination of proprietary and white-label open-source training datasets, environmental data from IoT sensors and internet data, and insect behavioural data from live video camera farm footage.
Entoverse CTO and Co-Founder Dr Dmitry Mikhailov says "ChatGPT and now GPT-4 have proven the potential of AI in changing the way we approach the problems of today.
He explains that one of the problems Entoverse is solving is improving the efficiency of insect egg production and larvae rearing for maximum protein mass output. This cannot be done with manual farm management alone.
"This is why we have built Entoverse, which automates the farming process and improves yields. FarmGPT will now take this automatization one step further, and we are excited to witness how it will transform current approaches to insect farming for better output.” Dmitry Mikhailov
FarmGPT is being prepared to be tested at CricketOne, a cricket farming company based in Vietnam. The trials will study the efficacy of GPT-driven insect farming processes and how it will benefit farmers. Future applications of FarmGPT potentially include the breeding of other species in closed, controlled environments, such as bees, mealworms, shrimp and fish.