Bcu.Cict.CmsApi.Entities.Content.Corporate.Content.MediaItem?.AlternativeText

Smart Greenhouse

Team Error 404

Teammates Palvir, George, Jack, Harrison, and Navid are Error 404 – a group comprising of BSc (Hons) Computer Networks and Security students and BSc (Hons) Computer Science students. Find out how they plan to create a smart greenhouse to help solve world-wide issues such as hunger and sustainable food production.  

Give us a brief overview of your project

Our project is a smart greenhouse management system using artificial intelligence (AI). We take data from inside the greenhouse such as temperature and plant water levels and keep them at the ideal level using our AI system.

What made you want to develop this idea?

This idea formed from wanting to develop a system that was practical but related to something we enjoy. Some of our team members are interested in gardening, so we wanted to create a greenhouse that can provide a stable source of vegetables all year round, in all conditions.

Why is there a need for what you’re creating

Access to suitable and locally sourced food can ultimately play a part in solving world hunger. When communities have access to fresh, nutritious, and locally grown food, it can improve their health and well-being. By promoting and investing in local food systems, we can reduce dependence on global supply chains and ensure that everyone has access to healthy food. It is an essential step towards building a more sustainable and equitable food system.

Can you explain about the kind of equipment you are using, and how you are creating the physical product?

We are using a Raspberry Pi-like, small, low-power computer to interface an Arduino board which is a small electronic controller board used to interact with our sensors and outputs. We are using a temperature sensor and a heater to regulate the heat inside the greenhouse, a humidity sensor and a fan to control how humid it is inside the greenhouse. We have a water pump and moisture sensor to ensure the plant is always kept in an optimal range.

Finally, we have a light sensor and lights so the product can work all year round. To control the optimal levels, we have created our own AI, which takes lots of existing data to find relationships enabling us to create the ideal conditions for a plant to grow. We have a web app using NextJS so you can view the data coming from the monitoring system on your phone or computer. It uses a custom Django web API so the data can be sent to the cloud if you operate a more extensive commercial environment. To visualise the data, we are using Grafana and Prometheus metrics monitoring systems. We have custom created this setup to make it as modular and enterprise-grade as possible, using enterprise-standard tools.

What challenges have you had to overcome?

We have run into and overcome many hurdles to complete our project; it hasn't been all smooth sailing. As we are using a similar system and not an actual Raspberry Pi, we ran into issues getting the code for interacting with the sensors to work. We had to edit the code and modify the operating system to get it to function. We also made many different prints and cuts to make the physical greenhouse stable. The electronics we used were also complicated to wire up as we had many different voltages to try and manage within the same system.

Where do you envision the future of this project?

We envision this project becoming a turning point for intelligent greenhouses, enabling small-scale farms, hobbyists and even enterprise operations to operate using the same underlying data. This will help push us towards having the best data to grow crops that can help eradicate world hunger nutritiously.