Vertical Farm Project - 2024-2025

Blogged by Andy Yeh

Prototype 1 (2024)

Prototype 2 (2025-2026)

How this got started - introduction

On 9 March 2023, I took a group of PNG teachers to visit Kelvin Grove State College (KGSC) and see their indigenous garden and aquaponic vertical farm. The PNG teachers are studying in my unit for a school-industry partnership STEM inquiry experience. The aquaponic vertical farm is one example of an outcome from school-industry partnership between KGSC and AUG (Australian Urban Growers). I have seen some hydroponic and aquaponic gardens (home-based) before, but this one is the first time I visited that is done by an industry, collaborating with a school. The farm is still small in scale but with its vertical advantage, it can yield good quantity of fresh produce.

I met Christopher Fulloon (founder of AUG) during this visit, and since then started to conceive a small movable vertical farm project for STEM education at QUT. There were some commercial product (e.g., Airgarden, sad to know they went in hibernation mode from 2025. Their website has very good curriculum resources) to work on, but in order to experience more authentic collaboration with industry, I chose to work with AUG to design one.

Initial discussion and preparation - November 2023 - July 2024

On 17 November 2023, I started an inquiry to Chris regarding building a small vertical farm outside of the STEM room (KG-B428b). My initial idea is to have a small movable vertical farm with one or two cylinders that has auto-irrigation and environmental monitoring (data logging with IoT, such as this indoor IoT pot plant project) for STEM studies. After a few conversations with Chris, and discussions about its size (e.g., number of cylinders), type (e.g., hydroponic or aquaponic), lighting requirement and location, water and electricity requirement, automation settings, type of vegetables, cost and on-going maintenance, and safety/risk assessment etc., a novel design of a rolling vertical farm started to take shape.

Initial design brief/quote

Using Fusion 360, Chris designed a Rolling Vertical Farm in 3D. This is even before any funding/payment so I really appreciate Chris's effort in doing this. With his experience in implementing vertical farms (e.g., the aquaponic vertical farm at KGSC), this Rolling Vertical Farm (RVF) looks quite promising. The size of RVF has become larger than originally planned, and so is the cost of building this farm. Thanks to my Head of School (Prof. Terri Bourke)'s support, we were able to secure some fund to build this RVF.

Location selection

The consideration of location was concurrent and a key part of the design. Due to its needs to access sunlight, water, electricity (for a pump and timer) and Internet connection (for IoT sensing and controls), the originally planned location outside of KG-B428b could not satisfy the requirement. After inspecting around the STEM room in B Block, the Level 2 courtyard right below KG-B428b was chosen to place the Rolling Vertical Garden.

This location has access to water and seems to have sufficient sun light. However, there is no power point at this location, so a request for a power point was sent to facility management. Thanks to Leigh Burgess, QUT's Sustainability Manager at Facility Management who provided experience on her Airgarden project at QUT since 2023. Thanks also to Naomie Buttini and Fabio Ribeiro from QUT Facility Management who helped inspect and approve this location for the Rolling Vertical Farm. Thanks to Peter Croft, two power points were then installed in July and the location is ready to welcome the RVF.

This initial discussion and preparation stage has very rich integrated STEM as well as project management skills. The STEM knowledge and skills for designing the first prototype of RVF would need another report/article to unpack.

Making, Installing and Running the Rolling Vertical Farm - August - December 2024

Engineering process for the making of RVF

When conceptualising this RVF in first half of 2023, Chris has started cycles of design thinking processes, to engineering the system (the RVF as a system) and its parts. I visited Chris's workshop where he empathised, ideated, prototyped, fabricated, and tested different components of the RVF. I took some photos of his workshop (right below) and Chris also shared some of his design (left below).

Put parts together for a working RVF

If considering the RVF a system (for sustainable growing of vegetables), then we can apply systems thinking to identify elements and their functions and connections in this system. The main elements are:

A water tank - provision of water and nutrients, and supporting the six planting cylinders standing up-right in the tank. The tank also has 4 wheels so the RVF could be moved around even the tank capacity + other parts weigh some hundred kilograms. The task also has an overflow mechanism to prevent too much water fill-in the tank.
Six planting cylinders - These are 90mm PVC pipes at about 1800mm in length. Each has some 20-30 holes/pods with an inserted platform to hold a plant. The bottom of the cylinders submerged into the tank water and it has a cap on top with small water pipes that mists water to flow from tank water to top of cylinder then down again to the water tank.
An industrial grade 12V pump and a Solenoid, connecting a water grid to supply water in this RVF system.
An electrical box to hold a smart wifi powerboard. This smart wifi powerboard is controlled by an App, which can set automation routine for pumping water.

Finally, the first prototype of this Rolling Vertical Farm was delivered on 26 August 2024. Chris and Steve transported all components from their workshop to QUT and installed at the planned location (see images on the right).

Start growing with more issues

The installation encountered some minor issues. For example, the cylinders were glued (industrial grade) to the tank but still came loose. That was fixed later so the first batch of vegetables/herbs were planted in the following week. The auto-irrigation system was set, which included the smart powerboard connecting to the Makerspace Wifi (at Level 4 above this RVF location) and automation scenario and timer to turn on and off the pump. At the first few days, this farm seemed to work well, but then more issues emerged.

The tank after filling in water seemed to deform its shape to cause misalignment of the six cylinders. The water pump also behaved abnormally and did not follow the timer and automation rules set in the APP.

Full initial setup with plants (2 September 2024)

The six cylinders did not stand up-right due to the deformation of tank and gluing issue.

Plants were not growing well due to abnormal water pumping. (30 September 2024).

Cycles of problem-solving - issue - debug - solution

To solve the cylinder up-right problem, a top capping (hat) was devised to help equally spacing out the cylinders as well as to improve the water misting.

The water pumping issue has been through many cycles of debugging, including airlock in the pump, solenoid, timer and time scheduling, weak wifi signal, and ultimately, the automation App (Grid Connect) issue.

Chris checking airlock in the pump

Checking solenoid.

Checking pump with no leak

A Wifi access point (AP) is required for the smart power board to operate. To begin with, we connect the power board to the independent Wifi router in the Makerspace. However, the connection was intermittent due to weak Wifi signal and this caused the timer/watering not operating as scheduled in the automation settings. To boost the Wifi signal, an indoor Wifi extender (RE205) is used but still could not produce good range and signal. Then with Chris's recommendation (he has experience/expertise in this as an industry expert), an outdoor Wifi extender (EAP225) was used. This did resolve the Wifi signal issue. For more details about the Wifi products and internet access for IoT projects, please visit this Internet for makerspace page.

During November-December, the farm is growing well some fresh vegetables and herbs. However, more learning and issues emerged.

The Wifi signal now is strong enough for stable internet connection. However, the irrigation automation can still act abnormally. After some more testing in the automation App, Chris realised that the problem originated from the cloud service of the App, in which the automation schedule, when interrupted (by the cloud service itself), could end up with multiple concurrent executions (or reentries) of the same schedule, and therefore, the pump could be in chaos busy pumping and sometimes not stopping or starting.

On reflection, this is another evidence of what expertise an industry could bring into the STEM classroom and schools, not just to implement industrial grade (as opposed to classroom/lab making) product, but also to debug and problem-solve.

On the other hand, the term "industry" usually imply reliable products and solutions. The industrial grade pump for example, is a good quality outdoor pump, and the rolling tank is well constructed. However, the commercial product smart power board has failed to produce the required results for this RVF project. Further, this industry designed RVF is not without flaws. For example:

the potting holes in the cylinders could not hold water for long and thus the timer for watering needs to be set for every 10 minutes, then depending on the weather and types of vegetables, frequency needs adjustment.
The misting of water from top down may not distribute water adequately to all potting holes/plants.
The tank only has a wire mesh cover (allowing water cycle in the RVF system). Overtime, the tank starts to infest some insects (e.g., some wrigglers) and deposit soils at the bottom of the tank.

Due to various design issues, Chris decided to pack this RVF system back to his workshop for improvement in late December. The original plan was to bring it back in February 2025 but unfortunately, at this location, B Block was under renovation and construction from January - June 2025 and the level 2 courtyard was occupied with construction materials and not accessible, Therefore, this rolling vertical farm has to be suspended and come back in August 2025.

Suspension and redesign of RVF - January - June 2025

Luckily the RVF has been packed away before the construction. There is no access to this level 2 courtyard during this period of time.

During this period of time, Chris has designed another RVF, aiming to use own micro-controller (i.e., an ESP32 micro-controller) to implement the automation timer instead of using other's cloud system that we do not have much control over it. Chris also aims to remove the bulk water tank so the potting cylinder could pot more plants.

Installing version 2 prototype of RVF - July 2025

After a long wait, the new RVF finally appear with some major redesign. On 18th July, the parts of this RVF were brought back and assembled. This new RVF together with the previous RVF were introduced to the Master of Education students in my STEM Inquiry Experience unit (EUN637) but unfortunately, I had to travel overseas from late July to mid August so the RVF did not get fully assembled and operational.

During late August to September, B Block was having some roof cleaning, and the water access was taken away by the roof cleaner. We were waiting for clearance from facility management to use the water, and had to change plan to operate this RVF from late October 2025. Hopefully, we could start a new test and have one harvest in November.

First trial of prototype 2 - November 2025

Finally, the site was clear, a missing part was acquired, the IoT system (using Shelly Plus 1 Smart Relay and Shelly App) was setup, water access and power access were set, we are ready to run the first trial of this vertical farm. On Thursday 30th October and two Fridays 7th and 14th November, Chris and a group of Griffith students came to finish the final touch and added baby Basils to grow in this vertical farm.

Lesson learnt: To build the prototype on site (away from the Workshop) often identifies missing parts and tools so it will take longer to build and make the RVF ready.

Chris is the mentor and supervisor for a group of 5-6 Environmental Science students from Griffith University doing their work integrated learning (WIL). That is another good example of partnership between university and industry.

And on 14th November, this prototype 2 of RVF finally began its first trial. There are six vertical cylinders and each has 33 holes to grow a basil, that comes to a total of 198 plants of basils in this vertical farm.

The watering timer was set for 10 seconds in every 10 minutes. The control mechanism was done by two smart relays, one for a pump, and another one for a solenoid, which does the actual control/trigger of the pump. The pump, when activated, will pump water from the middle container all the way up to the top and inside of each cyliner. The water then drip down to supply each plants in the pot holes. Water dripped down will be collected at the bottom dish and recycled back to the under pipe and middle containers. The timer by Shelly's smart relay seems to work well. But it was expected that the timing (e.g., 10 seconds for every 10 minutes) will need adjustment depending on weather and seasons.

This was an exciting day and we can't wait to see how the basils could grow or not in this vertical farming system.

The following Monday was a very hot day, with outdoor temperature soared to above 35 degrees in Brisbane area. We only remotely monitored the timer to ensure the planning watering was on schedule. We did not visit this farm until Wednesday, Chris found that almost half of basils did not get sufficient water and have dried. Chris then adjusted timer to increase water supply. Then on Friday 21st November, when visiting the farm, we found that every pot hole did receive good amount of water, but most plants have withered. Other issues emerged.

Chris explained that he has not yet implement suitable water chemistry for this farm. Key water chemistry for this hydroponic farm include:

pH Levels (5.5 - 6.5): The most critical factor is maintaining a slightly acidic pH between 5.5 and 6.5, which allows for optimal nutrient uptake. A pH outside this range can lead to nutrient "lockout," causing deficiency even if nutrients are present.
Electrical Conductivity (EC/TDS): EC measures the concentration of dissolved salts (nutrients) in the water. For most plants, the EC should be between 0.8 and 2.5 mS/cm, or a total dissolved solids (TDS) reading of 200–300 PPM, though this varies by crop.
Water Temperature (18-22°C): The nutrient solution should be kept between 18°C and 25°C (64°F - 77°F). Warmer water holds less oxygen, while cooler water holds more.
Dissolved Oxygen: Adequate oxygen is vital for root health. The water should be well-aerated, with an Oxidation-Reduction Potential (ORP) of 300–500mV to prevent pathogenic growth.
Water Hardness and Alkalinity: Water should not be too "hard" (over 150 ppm of calcium/magnesium), as it can clog equipment and cause pH instability. Alkalinity (bicarbonate levels) should be managed to prevent rapid, uncontrollable pH increases.

At this stage, Chris has developed a good idea for watering with the Shelly IoT timer/relay, and has planned to replace all plants in early December for the second trial. In the image carousel on the left, we also noticed that the water tank needs to refill manually every week. There seemed to have some water leaking and/or recycling problem.

On Friday 28 November, we confirmed that there must have some leaking or the cup at the bottom of each cylinder did not catch and recycle the drip-down water effectively. The pictures below show that the water spilled on the ground and again two empty tanks to refill. There is a need to install some water refill automation for this system. On this day, I also received a weather station ordered for this farm and started to plan how and where to install.

Water spilled on the floor

Two empty tanks in a week

Refill water from one will fill up both.

On Friday 5th December, we started the second trial. Chris has some helpers (apprentices) from Griffith University. This day marked another new implementation/step forward for this vertical farm. Some key changes/improvements are:

The farm has installed more water pipe lines connecting the rectangular base PVC frame and the two water tanks. This will increase the water circulation and ensure the water levels in the two tanks and the PVC frame are balanced.
A water refill mechanism uses a buoyant float connected to a valve to automatically manage water levels was installed. When the water level drops, the float lowers, opening the valve to fill the container; when it rises, the float closes the valve to prevent overflow. With this mechanism, we will leave the water tap on.
Water chemistry and nutrient were adjusted and added with the product ISN A+B BLOOM. There is rich S.T.E.M knowledge in this and great potential for integrated STEM. Chris instructed his Griffith apprentices to measure and mix the ISN A and B products in certain amount and ratio for the farm. This has factored into account the type of plants, the capacity of the two water tanks. TDS and EC were measured using an instrument after the nutrient (ISN A+B Bloom) was added in the tank. I did not record the readings nor did I understand what would be good readings in this farm system. However, I can imagine what students could learn about STEM and engage with STEM industries to solve meaningful real-world problems.

At the end of the day, the farm is put in place with new fresh baby bassils and lettuces. Unfortunately, I have one more week at work for 2025 and will then take annual leave. Next week, I could hopefully see some growth of the plants and must have the weather station installed at this location.

8-12 December was my last week at work for 2025. I was pleased to see the plants actually grew bigger and with the help of Facility Management (Naomie Buttini, Ross Benson, Heath Chaney), the weather station was up and running on Friday 12th December 2025.

The rolling vertical farm was in opertaion until late December then was moved to Kelvin Grove State College where Chris has collaborated with the school and established a much bigger vertical farm. The rolling vertical farm was moved over due to the closure of University during Christmas holiday. As a "Rolling" Vertical Farm, it is moveable and could be better looked after at the school by Chris during holidays. On 9th December, Chris showed me the produce (images below) from KGSC's vertical farm, and assured that this rolling vertical farm will be able to achieve the same.

That all for 2025. We look forward to the next journey in 2026.

Page updated

Google Sites

Report abuse