Premisse

A chatbot that helps you as a plant owner with the care of your plant. And in the long run helps you to get green fingers.

Synopsis

A subject that I actually deal with on a daily basis, keeping my houseplants alive. Even though I try so hard, it never seems to be enough. In the end I’m left with hanging, dried-out tasks and leaves and an inferiority complex. Therefore, state of the art in technology is the perfect opportunity to tackle this problem and get involved in this topic, because I can’t be the only one!

Project context

Research among 2,000 millennials, aged 25-39, shows that seven out of ten millennials call themselves ‘plant parents’. Despite the high demand for houseplants among millennials, they underestimate the care of these plants. Indeed, 67% of millennials say that caring for plants is more of a challenge than they had anticipated. (2020)

The survey also looked at the attitudes and behaviour of millennials towards plants and found that 22% are worried about owning a plant because they have accidentally killed one in the past. (2020)

“The average plant parent has killed seven plants”

The biggest challenge in plant care was how much sunlight the plant needed. With the challenges of how much water does my plant need, should it live inside or outside and how do you prune a plant. (Baas, 2019)

The biggest and also most difficult lesson that three out of ten respondents learned was how easy it can be to accidentally kill your plant like a good ‘plant parent’. This causes anxiety, as three out of five ‘plant parents’ often worry about whether the plant gets enough sunlight, while another 56% worry about whether the plant has enough water.

Yet the plant parents indicate that being a plant owner is addictive. Indeed, 41% of respondents say they will expand their plant collection in the coming year. Time to fix this problem!

The following main and sub questions have been formulated to guide and support the research:

Main question: How can a digital application help people to take better care of their plants and motivate them in the long run?

Sub-questions:
Who is the target group?
Where do things go wrong? (Most common mistakes when taking care of a plant)
What are the wishes of the target group?
Who are the competitors and what do they do?

Research

Who is the target audience?

Research shows that the older the plant owners, the easier they found it to look after the plants. The millennials between 25 and 36 years old experience the most problems while taking care of plants. That is why we have chosen to focus and test on this target group.

Within the target group, you have two subgroups: novice and experienced plant owners. The experienced plant owners have a fixed rhythm with fixed days for watering the plants. They also feel with their finger in the earth to see how much water the plant needs. While a beginner finds watering the most difficult.

Where do things go wrong? (Most common mistakes when taking care of a plant)

Besides knowing who the target group is, it is important to know what problems they experience when taking care of plants. By means of a literature review, insights came to light that I wanted to confirm or invalidate. Twelve people from the target group were interviewed for this purpose.

Most common mistakes when taking care of plants:

• Too much water (or too little)
• Not taking the humidity into account
• Removing unhealthy leaves
• Too much light

(Lembo, 2018)

The twelve interviews showed that the plant owners find watering difficult. Quantity plays a major role in this. They often either give too much or too little water. Half of the interviewees indicated that they have no fixed day for watering the plants. This was particularly true of new plant owners.

This corresponds to what came out of the research. It also showed that watering is the biggest problem among millennials.

Another problem is that they don’t look at the humidity. For plants, it is important not only to water the plant, but also to spray the leaves wet from time to time. However, the target group is not aware of this. After they did not start talking about it, it was presented, 10 of the 12 persons from the target group did not know that this was necessary. (Baas, 2019)

What are the wishes of the target group?

I also asked the target group what they thought they needed when looking after a plant. Various points came up:

• An overview of the different needs of their plants
• Setting reminders for these needs
• Confirmation that they are doing it right
• Wanting to know what is wrong with the plant (if the leaves change colour)

Who are the competitors and what do they do?

Blossom
The application blossom identifies plants and gives the user ho-to’s and useful tips. In addition, the user receives reminders to water the plant on time without drowning it. The application does not use sensors to monitor the plants.

Plantsome
In the application, you search for your houseplant. Then you indicate the size and how much sun the plant gets per day. The application calculates when and how much water and nutrition the plant needs, using the local weather report. The application then notifies you that the plant needs water and how much water you should give it.

Happy plant
Watering plants in a playful way. By using gamification: watering streaks, naming your plants like babies, plant-selfies, and time-lapse videos to watch your plants grow.

Creative process

Concepting

Based on the information gathered, it was time for a brainstorming session. The aim of the brainstorming session was to come up with a concept that could be used during the further process. The ‘crazy 8s’ technique was used to come up with three concepts. This is a quick exercise to come up with eight variations on an idea within eight minutes. (Knapp et al., 2016)

Concept 1: Personal chatbot that helps you take care of your plants.
Concept 2: Monitoring of my plants in a greenhouse linked to an application
Concept 3: An application that gives advice on watering your plant based on local weather forecasts.

By questioning the target group, it emerged that they need personal advice. They want to be able to ask questions about their own plants. The majority of the interviewees therefore opted for concept 1. However, they also thought that monitoring the plants was a good idea, because you could then see more specifically what a plant wanted.

First prototype

I started by sketching my ideas on paper. These soon became realistic after the chat bot also got its own name. This was one of the requirements of the target group, that the chatbot should feel personal, so that they had the idea that they were getting advice from a plant expert. (Baas, 2019)

The name Carry, derived from taking care, was chosen. After a short brainstorming session, I let the target group choose from three names:

• Remember me
• Carry
• Plant-to-know

The twelve people in the target group unanimously chose Carry. Then it was time to develop the chatbot further. Pens and paper were used for this. This allowed the dialogue and ideas to be worked out quickly.

Second prototype

These sketches were then further developed in the prototyping tool Figma so that they could actually be presented and tested with the target group. The elaborated dialogue focuses on watering. By means of icons, the user can answer the questions of plantbot Carry.

I tested this prototype with six people from the target group by means of a task-oriented user test. The task within this user test was to water plant Paul. Plant Paul’s large felt dry, and you wanted to set a reminder to water Paul every Tuesday at 21:00. (Knapp et al., 2016)

Insights:

• Users want to be able to type their own answers
• More interaction with Carry
• More specific information about the plant

Third prototype

During the test, the target group indicated that they wanted more interaction with Carry and to be able to type answers themselves. Therefore, I chose to use power visual agents from microsoft to create a new version of Carry. I had never worked with this programme before, but soon I had realised the first version.

To see if I had chosen the right trigger phrases and words, I did a quick test with three people from the target group. This involved the same tasks as the test of the second prototype. I added the missing trigger sentences and then did a final test with all twelve persons from the target group.

 

Insights:

• Still uncertainty to see if they are doing well, even though the description is already clearer.
• The humanidity component is now not included at all, while it did come up during the literature review.

Arduino prototype

Sensors: Most important for monitoring the plants environment is a soil moisture sensor.

I connected my moisture sensor to my coarse beginner kit for Arduino. I put it in the A0 port. Then I connected the arduino kit to my laptop through the USB.

After watching some tutorials and experimenting with finding the right code. I finally typed the right code into the IDE and the moisture of my basil plant appeared in my terminal. (Kamaruddin et al., 2019)

 

Conclusion & reflection

It was great fun to be able to experiment with technology during this project. I learned a lot about plants, taking care of them, making a prototype in Figma, working with chatbots and Arduino. It is actually possible to help the user to take care of a plant by means of a chatbot. The addition of an Arduino soil moisture sensor ensures that this advice can be even more accurate.

In the future, I would like to link these two so that I can also test this version with the user. I would then spread this test over a longer period of time so that we can really look at the long term.But for now, it was already incredibly interesting to involve the user in the process. I would also like to experiment with rasberry pie in the future, because then you have even more possibilities.

 

References