How might we use AI to build a community around Urban Farming?
What is FarmVision?
FarmVision is a family of products that make your farming journey delightful by managing segments that cause pain, providing crop insights and connecting communities, all while growing a story with your plants.
Problem area: There is an increasing demand for Urban Farming which started as an antithetical to traditional agricultural practices that is one of the major cause of carbon emissions today. Although automation and Artificial intelligence is prevalent in traditional agriculture, the use of AI products in Urban farming is very nascent. It is difficult to replicate AI's use from traditional to Urban farming, as they run on very different ideologies, one relying on profit and the other on community. This market gap gave us a lot to work with in the field of AI for Urban farming.
Solution: An AI assistant, both wearable and Mobile app, that helps the farming communities to plan, manage and harvest the crops for a cycle. This is done by monitoring plant development through image recognition all while growing a story with the plants.
Timeline: 4 months Focus: Design for AI Team: Elizabeth, Jingyi, Ting, Shashank, Harika
My Role: As a researcher I was responsible for doing the secondary research and also conducting Primary research with SME's and collecting Work artifacts for visual presentation. As a designer I was given the task to design the first iteration of the Glass facing experience. Working with the team I also helped in the process of telling our story through an interactive and engaging presentation.
How do we use AI to build a community around farming rather than mindless automation?
Through our secondary research we identified problem areas were could potentially create an impact through design intervention that focuses on Artificial Intelligence.
Analyze soil conditions
AR experience to train novice farmers
Suggest how and where to plant the seeds in the community farm
Aid in crop growth by managing and protecting it
Gamification of the process
Timely alerts to show mature crops during harvesting
Smart distribution of food to the community
Our next steps were to fix the scope for our product by understanding who we are creating the product for (User Frame), what the product is going to be (Business Frame) and what problem area are we solving (Problem Frame). This exercise was useful in getting everyone onto the same page on the direction the product would take and carve our product features more intuitively.
We listed down all the users groups and all the activities involved in the farming process. We then marked down areas we wanted to concentrate on and the potential users for that operation.
This helped us guide our interview scripts and recruitments.
Our next step was to identify organizations that would potentially implement our design and the scale of our solution.
These frames were for our initial ideas and is flexible to evolve as the research progressed.
Lastly we combined both the frames to arrive at a Problem frame where we understood the end user, the features, the organization that would be using the product and a tentative name to describe this solution.
Reviewing the methods already in existence, we tried to explore areas that haven't been touched upon in an Urban farming context
Plan your garden
Crop health monitoring
Done and dusted ideas
What we want to explore
We dived in to market research, understanding work that is already done by others to see where we can situate our product in the market. This also made us aware of the work that is being done in the field of urban farming.
The five pointer scale helped us see visually, the areas that are lacking in the existing products in the market.
We held sickles and ploughs and deep dived into the field (pun intended) of Urban Farming. We visited 3 varied locations to understand our users, their journey and if urban farming is what we assumed it is. We also interviewed 2 Subject matter experts, 4 volunteers and surveyed people who are interested but don't practice farming completing our Primary research.
Urban roots was a Non profit whose main function is to "empower, engage and nourish" the lives of young people through farming. Compared to 3 farms we looked into this was by far the most conventional urban farm. it uses traditional sustainable practices where the main workforce are volunteers. You can see our team volunteer in the images to the right.
Microfarm is a University of Texas initiative that is a student led organic farm where students or faculty can rent a plot to grow their veggies in ways that they are comfortable with. The user group for microfarm are novice inexperienced users who want to experiment and want to know exactly where their produce comes from.
The sustainability department at UT grows produce that they then supply to the Food and Dining division, all using sustainable and alternative practices. This was by far the most experimental farm of the three, with the use of aeroponics, hydroponics and IPM method.
Co-Creating Journey Maps
Office of Sustainability
Office of Sustainability
We have seen in all our field studies that people depended on physical copies to manage and plan farms. When probed further into why they had to rely on them they just shrugged and said "It works, I guess"
An Excel sheet 'plan' of a Garden that is managed by the Department of Sustainability. This is how they manage and organized their crops.
Mural Board Externalization
Following IBM's design research methods, the team got together to make a number of Mural Boards to codify the data that is collected
Who are we Designing for?
Manage volunteers and farm
Tell a story of the produce to leverage the Farm in markets
Get more people involved
Teaching new volunteers about the procedures in the farm
Too much work during harvest season
Failing to keep up with the people reporting to him Unfamiliarity within the building codes
Volunteers / Farming enthusiasts
Produce own food for consumption
Be at the farm whenever possible
Understand how to take care of the plant that is showing signs of trouble.
Make friends through the farm
Uncertainty within the management
Using the user journey and understanding their pain points, we sketched scenarios where our product has a potential to be used by the users. This made us slowly develop our ideas towards something more concrete.
Jackie get's a notification about her plants through the app. She plans her routine, ticking all that she wants to do today. This is updated into the Farm manager's database where she keeps track of all her volunteers.
On arrival at the farm, Jackie picks up a glass. Through iris scanner the system knows it is her and provides the necessary information to finish her tasks
Disease scanning, soil prep, optimal field prep and weather conditions are calibrated to help with the tasks.
After harvesting, the produce is displayed in the farmers market on the boards on on the mobile application through a QR Code
Jackie just got out of bed, trying to plan her Sunday, as she gets a notification from FarmVision. She is instantly hooked now and the cycling process begins.
She goes in to check the app and is now provided with a list of tasks that she could do if she visits the farm today. She can now go ahead and choose the tasks she likes and "Confirms"
The farm manager's database is instantly updated with a list of volunteers coming today and the tasks that they are able to finish. This keeps them on track to efficiently manage the farm without worrying about the menial tasks.
At the Farm
Volunteers pick up their glasses during the orientation session. The Glass which is called FarmVision, is an interface which helps the volunteers work with their hands and sets them on tasks.
FarmVision is connected to a variety of databases like weather, a farmers almanac, image recognition for diseases, and sensors that are placed at the edge of the rows that pick up on the plants health.
FarmVision reads your retina/voice and welcomes you with a personalized greeting. When asked
"Hey FarmVision what are my tasks for today?"
It provides with the list you already selected back home.
The involvement of the AI glass is put to a minimum to enjoy the process of Urban Farming.
FarmVision also has some Easter eggs, where at times you have plants interact with you through AR technology. This makes the journey fun and gives a personality to the plants, which could then be developed to make our story.
When the plants are finally harvested, they are then taken to farm stands to be sold. A QR code is stuck on every bunch, which the customers can scan to get information about the plants. The data for this story is captured by AI to present a story about:
- Origin Story
- People involved in growing them
- Personality trait of the plant
- It's growth pattern
- What practices were used to grow the plant
- Funny anecdotal stories
User research forms the backbone of good design
Every design has a purpose and that is to reach the intended audience effectively. We spent a lot of time doing extensive research and understanding our users and the process in depth. The knowledge base we created proved to be extremely useful throughout the design journey, allowing us to make informed design decisions at each step
The first design is almost always never a perfect solution. Iteration helps, and is essential
Explorations and re-iterations based and research and user testing is what plays a crucial part in determining how well the product would serve its users. Working and changing the design direction is better than letting the process stagnate.
Design is at its best when ideas are discussed freely
While working in a team, creating a productive and idea-oriented environment is crucial. It is important to be sympathetic and take constructive critique of all ideas, even if it is yours. When individuals are free to express, the ideas that sprout from the smallest of comments can make a huge impact on the product.
Defining the problem, documenting, and having clear goals matter
While working on a time-bound problem, and that is always, it is useful to have an idea of the project timeline. Timeline updates according to product evolution, but having an overview in mind gives context to each step. Documenting design decisions help maintain a progressive flow, and journey efficient and fruitful.