Warning: Only the first byte will be assigned to the string offset in /usr/www/users/integjpmre/wp-content/plugins/responsive-menu/v4.0.0/inc/helpers/custom-functions.php on line 218 Integrove

How to build an autonomous garden with AI and IoT

autonomous garden

Brett Hudson & Angela De Loureiro
18 August 2022

Who would’ve thought that growing a few peppers for the office would require a team of technology experts? I mean, haven’t plants existed for millions of years without us? That was the DevCrop team’s thoughts when we suggested this project. We theorized a project where we could combine our technological savvy with an industry not well known for its information innovation: agriculture. Mechanically speaking, commercial agriculture has seen some epic advances in technology within the last 50 years; have a look at the John Deere website to see some of the incredible machinery they’ve produced.

It seems that the planting and harvesting side of things is well taken care of, however, in the crop monitoring space, there’s been a bit of a lag. Not looking to compete with the ‘big guns’ just yet, we thought of placing ourselves right there in the middle, monitoring the inputs and environmental conditions of crops after planting and before harvesting. Essentially, we’re building an autonomous garden.

Building and autonomous garden – Project DevCrops

Project DevCrops is a hydroponics initiative developed inside the Integrove Innovation Lab with the aim of figuring out a way to automate and enhance the crop lifecycle. Hydroponics is the process of growing plants using nutrient-rich water rather than soil. The main idea behind hydroponics is the ability to control the inputs given to the crops and to reduce water consumption, pollution and chemical use associated with traditional agriculture. Hydroponic growing methods remove the need for expensive and harmful pesticides, fumigants, and herbicides, saving on input costs while ultimately producing higher quality yields. The main drawback has been the initial capital required to set up a commercial hydroponics system.

Building an autonomous garden – our design

There are several different systems available for growing plants hydroponically, we chose the following design as it was the most suitable for our monitoring activities:

autonomous garden

We initially wanted to measure some simple metrics like water temperature, pH level and the level of nutrients in the water. We soon realized that we may have stumbled onto a ‘gold mine’ of potential data collection activities in the process. Project DevCrops is three months old, and we are already incorporating IoT devices, micro-controllers, cloud storage and dashboards to monitor and analyse our autonomous garden.

Why build an autonomous garden?

At this point you’d be forgiven for thinking that this seems like overkill just to keep a few plants alive; and who grows food like that anyway? Well, the Dutch do, and they are beginning to feed the world!

The Netherlands is currently the second largest exporter of food by value globally. Second in exports only to the USA and the size of West Virginia. They’ve been leading the way in smart, sustainable agriculture and have mastered the development of ultra-efficient, semi-automated greenhouses, which require a fraction of the water and chemicals than traditional agriculture. The Dutch are obsessed with drawing out the highest yield per m2 possible and have harnessed all the technologies at their disposal to achieve this, including the IoT devices, micro-controllers, and cloud technologies that we are using for Project DevCrops.

Due to the nature of the project and the requirements for intervention, we have designed a solution around IoT devices and the ability for near real-time streaming.

How we can help you

Like anything that is true to our Integrove way, we build bespoke solutions. Applying Scaled Agile, we designed a solution around IoT devices and the ability for near real-time streaming. Operationally, we are making use of an ESP 32 microcontroller to manage our IoT devices, and we layer data science and data analytics capabilities onto the captured data for further analysis. Here is our design:

This solution is designed in a way that will accommodate how we expect this project to grow and change over the next few months. The final goal is to build a fully automated system where devices can adjust the variables as needed. Keep an eye out for our next blog where we let you know how our plants are doing!

Contact Us