-Millennials and post-millennials often grow up disconnected from food preparation despite the indicator of selfpreparing meals being the strongest single indicator for longterm health. Up to now, our higher education system had struggled with implementing projects that help students overcome this disconnection and thereby teach them holistic approaches spanning multiple disciplines, which could also benefit their future careers. We conducted a multidisciplinary pilot between computer science and hospitality management within a senior design project class and service learning capstone class on food sustainability that implemented two versions of a resilient smart garden to compare their yield to hand-watered growing boxes. In this paper, we report on the set-up, discuss lessons learned from the pilot, and project future scenarios leveraging the sustainability transformation mindset principles that support transitioning towards teaching sustainable livelihood with the support of ICT.
Millennials and post millennials often grow up disconnected
from food sourcing and preparation despite the indicator of
self-preparing meals being the strongest single indicator for
long-term health [
To reduce an individual’s outdoor water usage, researchers
and practitioners have developed automated watering
systems [
Therefore, the purpose of this project is to engage
millennials in growing food and thereby using this fresh whole food
at home as well as to address the issue of food security on
campus. The long-term goal of the project is to address food
security issues related to lack of access and utilization. In the
CSU system, one in five students do not have steady food
access, creating barriers to the ability to learn [
The contribution of the paper at hand is that, based on a
preliminary prototype reported on in [
Our Resilient Smart Garden helps to ensure vegetable plants are not over or under watered. The aim is to find the water balance to grow the perfect vegetable. It also takes the mandatory watering out of the users hands and automates watering based on the moisture of the soil. The Smart Garden takes moisture and temperature readings to decide if the garden needs more water. These readings are stored in a database that is accessible online and can be accessed through PC or mobile smart phone. The data can be used to further research on the best watering method that fits for different kinds of plants.
The impact of our work is that the set-up, discuss lessons learned from the pilot, and project future scenarios leveraging the sustainability transformation mindset principles can support transitioning towards teaching sustainable livelihood with the support of information and communication technology (ICT).
A more in-depth treatment of related work for this project
is reported on in detail in [
Food security encompasses the ability of individuals,
households and communities to acquire food that is healthy,
sustainable, affordable, appropriate and accessible [
The pillars for food security indicate how well the system
is taking care of its constituents by assuring food, as a public
good, is accessible, available and utilizable by all citizens
equally. Food insecurity, a household-level economic and
social condition of limited or uncertain access to adequate food
(United States Department of Agriculture) [
Food security is a complex sustainable development issue,
linked to health through nutrition or malnutrition, but also
to sustainable economic development, environment, and trade.
Approximately 1 in 9 people on earth are food insecure. In
the United States this equates to 49 million Americans living
in food insecure households [
A study by Barthel and Isandahl (2012) found lessons
learned from different historical and cultural contexts (the
Classic Maya civilization and Byzantine Constantinople)
support that urban gardens, agriculture, and water management
contribute to long-term food security for people living in
cities [
There are a few commercial-off-the-shelf “COTS” systems
that are available in the market to help gardeners grow plants.
The search led to identifying Edyn Smart Garden System [
There are also a few electronic DIY projects are more
accessible with easily programmable single board
microcontrollers. Daniels [
The Guarduino project [
All of these projects have similar approaches to implementing an automatically watering garden. Our Resilient Smart Garden shares some characteristics to minimize water usage while maintaining a sustainable environment for the plants. The main difference is that we perform the moisture sensing on a plant-specific basis.
In [
Multidisciplinary research is highly valued by all funding agencies in theory, and in practice there are many hurdles that need to be conquered. However, the learning experience for both sides has been insightful and merits the effort. Multidisciplinary capstone projects are an easy introduction to conducting multi- and/or interdisciplinary research but, because of the higher number of involved people, require even more organizational overhead. We saw that overhead but still thought it was a good opportunity to try out the concept and then decide whether this should be made possible for students on a wider base or only in special cases.
Our long-term vision is to integrate this with permaculture
principles, where a garden built of plant guilds can foster
human independence from extraneous materials and promises
to deliver the highest harvest yield while making keeping the
grounds sustained [
Last but not least, we applied the transformation mindset
tool proposed by Samuel Mann in 2017 [
In the iteration presented in this paper, we added a multidisciplinary research component and teamed up with a hospitality management professor and her students to set up a comparative experiment to see whether the automated watering planters relying on the Arduino board constructions could keep up with the hand-watered planters by the hospitality management service learning team.
The senior design project course is a capstone course over
two semesters where students are in teams of three to six
and develop a product from scratch. In the first semester,
we usually follow a more traditional process of requirements
specification, design specification, test specification and
implementation. In the second semester we move to an agile
model with several iterations. That way students are exposed to
both common paradigms. In the second semester, students are
allowed to work largely self-directed based on their previous
experience from the first semester. They report back weekly
and we hold reflective meetings to enhance their own analysis
skills and learn from how the project unfolds [
Students enrolled in a general education capstone course called “Exploring a Sustainable Food System” are required to complete 20 hours of service in the community. In this course, students address food justice in the community. The Resilient Smart Gardens was one of the projects the students could choose to encourage home gardening as a means to increase food access in communities. Three student leaders were identified and trained to organize the daily watering, maintenance and data collection of the project. The student leaders then trained, scheduled and managed the volunteers while reporting and consulting regularly with the Supervisor of the project.
The comparative experiment was set up to find out whether two planters using two different implementations of the resilient smart garden idea could achieve as much harvest as the two hand-watered comparison planters. To be able to harvest after only two months, we planted two specific types of kale and romaine lettuce. It was run using a special soil developed for needing less water.
In the time line for the semester, the roles, tasks, and milestones were the following:
Roles: Supervisor for the computer science students was Birgit Penzenstadler, supervisor for the hospitality management students was Libby Gustin, supporting domain expert for the gardening with the special soil was Christian Anca.
Tasks: The computer science students got their implementations ready to deploy in the garden, the garden expert delivered the special soil and seedlings2, all teams planted and put the systems in place, the hospitality management students watered their planters and took readings of all water meters (see instructions in Fig. 1, 2), the computer science students monitored their systems, and we jointly harvested at the end of the semester.
Milestones: Planting and deployment on October 10th, harvest on December 7th.
2The semester was too short too pull the plants from seed and get all the way to harvest.
The evaluation compares the amount of harvest across the planters. Furthermore, we qualitatively explore the lessons learned.
VEGGIES
This section describes the steps we took — application of the transformation mindset, implementation and deployment, growing and monitoring, and harvest and results.
As Mann et al. [
Students analyzed the application of the Transformation Mindset principles before they went about the detailed design and the implementation. We also revisited the principles afterwards during the reflection phase.
The resilient smart garden is a low-key, hands-on version of growing food, useful for communities.
business as usual: The idea of growing their own food instead of choosing what to buy gives students more agency.
plinary project helps broaden the students’ horizon.
Equity and diversity over homogeneity: This is also one
of the principles of companion planting in
permaculture [
ering: Taking care of fellow students by donating to the food bank was appreciated.
rance: Growing food is empowering.
gave students new perspectives.
dents can choose to grow (at least part of) their own food.
perceive the opportunity as positive in the face of food deserts in the local drought-prone climate.
All setbacks in the project could jointly be overcome.
The computer science students had initial prototypes from the end of the first semester of their senior design project. However, taking those prototypes out in the wild required more effort as it turned out they didn’t run stable yet. Furthermore, challenges were to connect them to a wireless module for data transmission as well as getting them connected to solar power. In addition, the databases needed to be set up and connected. All of this took them several weeks at the beginning of the semester after the initial two weeks of getting everything back up and running after the summer break and designing their individual sprint plan for the semester. Figure 3 shows one of the two prototypes before deployment “in the wild”.
In the mean time, the supervisors prepared the remaining time line, purchased required hardware, and organized the delivery with the domain expert. The hospitality management students were briefed according to the protocols in Fig. 1 for the hand-watered boxes and Fig. 2 for the automatedly watered boxes.
Despite the challenges, we managed to deploy everything on October 10th with a few software updates over the following few days, see Fig. 4 for a close-up of box B2, and Fig. 5 for a closer look at box B1 with all other three boxes lined up behind and the sign asking to not tamper with out watering system in the front.
We started monitoring the systems and found a few glitches. It turned out that one of the systems was overwatering due to a not well fitting hose connector, see Fig. 7, so a puddle gathered next to the planter, see Fig. 6, that was pointed out to us by the service learning student volunteers. We fixed that, but this prototype (planter B2, Fig. 4) continued to slightly overwater despite the team repeatedly reducing the watering time. We received several emails from the student union sustainability representative over the course of the experiment asking us to make sure we are not wasting water. This box did end up using significantly more water than the other ones.
At the end of the experiment, the B1 automatedly watered planter used the same amount of water as the hand-watered planter A1 that had used less water (last readings from December 5th):
Hand-watered box A2: 280.93 gallons Automatedly watered box B1: 143.33 gallons
Automatedly watered box B2: 1022.25 gallons While having only two boxes per type does not allow to draw general conclusions, we see that there can be water wasted using either approach and that one of the automatedly watered boxes and one of the hand-watered boxes using almost the exact same amount of water (0.9 gallons difference) over the course of two months.
Box B1, which did well in terms of not overwatering, had a different problem though caused by the exact setting in the garden: There was a large tree overshadowing the box for a large share of the hours the sun was shining on the garden every day. Therefore, the plants did not grow as well.
We harvested all grown kale and lettuce on December 7th and took it to one of the kitchens in the Department of Family and Consumer Science to clean and weigh the vegetables. In Fig. 8, box B1 is being harvested by the computer science students. In Fig. 9, the entire team is standing in the hospitality management kitchen behind the cleaned produce.
The exact amounts of the harvest are listed in Table I.
Overall, the hand-watered boxes delivered more produce. The
number of lettuce heads and kale bunches is in the same
range for all boxes, which means no plants died off after
we planted the seedlings. There is a clear difference in how
well the produce grew though. Both the kale harvest and the
lettuce harvest amounted to significantly less weight in the
automatedly watered boxes. One of the reasons for the lower
amount of B1 could be that this box had less sunlight as
mentioned in the previous subsection. However, as box B2 did
not have that problem but achieved roughly the same amount
of harvest, we know that this is not the only and probably
not the strongest influencing factor. Box B2 did receive too
much water and as noted by Caetano et al. [
After each team member was allowed to take some of the produce for personal use, we donated the remaining large part of it to the local CSULB food bank3 for students on campus. Bringing the fresh produce to the campus food bank was not only rewarding to the students, it increased their awareness of the need for healthier food options at food banks as well as recipe suggestion and cooking skill support. As a result, all campus garden members are donating fresh produce to the food bank every 2-3 weeks. Hospitality Management students are developing simple recipes to use the food, and consulting with food bank patrons about cooking skills and utilization.
Around the harvest day, we sent an email to all student participants to ask for their observations and lessons learned: What were your major insights from working on the experiment? What were things you did not expect? What would you do different if you had to do it again from scratch? What else would you like us to know about your experience? We collected the answers and analyzed them to improve the project organization and management for a future replication.
There were some project management insights, e.g. This kind of project required contribution from different departments. Much like in the real world were your assigned a role in a project, however to continue and complete the project you 3https://www.asicsulb.org/corporate/discover/beach-pantry must collaborate and communicate with others who might not being working on the same part of the project.
Several students would have liked to have more in-person meetings across the disciplines, e.g. Another insight I gained was that not meeting with people in person weakens communication.
The hospitality management students reported insights on growing vegetables: I learned about how much water really goes into growing vegetables! and I learned a lot about how to water and how much watering and care it takes to have successful growth and maintain a garden.
The computer science students reported technical accomplishments, e.g. My major insights on the experiment were the extended capabilities of the Arduino and its ability to be integrated with other technologies like the ESP 8266 WIFI module and the SD card reader module. as well as The understanding of the data sheets and schematics of all the components really helped when trying to get the system to function properly. Having the option to solder and learning how to do so properly, allowed for easier customization of the system in comparison to using a shield for the Arduino. The B1 team tinkered with their system in creative ways to optimize the accuracy: Adding resistors to the sensor system helped in the voltage drop to provide a more accurate reading, yet due to the complexity of the current being impacted the readings weren’t as consistent.
There were also a few technical challenges, e.g. It’s also extremely difficult to use a board like the Arduino and try to connect to the school’s wifi because of the networks encryption protocols. Lastly, I learned it is very difficult to do weekly maintenance, which is probably why a lot of companies don’t periodically release updates on a weekly basis and more often do so on a bi-weekly or monthly status.
Trouble with the readings on the water meters was reported by several students, e.g. There were quite a few mishaps with the water meters. They were very finicky and created a lot of stress for the project. and One thing I did not expect was having to deal with the equipment not working properly.
About trying to get ready for deployment, one student reports: Originally when trying to control the water access using a solenoid valve, I had multiple attempts soldering together components to control the valve to open and close when prompted, which set back the ability to test the watering schedule.
Furthermore, some of the sensors had problems: We did not expect to have trouble with the data readings the sensors provided. When testing the functionality of the sensors with the
on the datasheets. When wiring the sensors to the WIFI modules for the data to be passed and stored on the Arduino wirelessly, the readings for the soil moisture and UV sensors were very inaccurate.
The student who tried to solve the over-watering problem of B2 explained: Some things I didn’t expect to occur was the constant over-watering. No matter how many times I messed around with the parameters on when to water the soil around our planter would remain moist. Even after days where the garden system didn’t water the soil around the planter never fully dried. I also didn’t expect to find out that it is impossible to connect to the school’s wifi. Online there’s all these claims where people have attempted and been able to do it. But, after attempting to connect using the same method others claimed to do online it still proved to not be possible.
Students would have wanted to up their gardening expertise ahead of time, e.g. I would also research more on garden maintenance and how to properly take care of a garden so that we could prevent over-watering. VI. DISCUSSION OF LIMITATIONS
Also, they’d want more interaction, e.g. Cooperate more We seem to have run into a few classic problems of ICT4S
with the other groups/departments involved. and I would in student and researcher projects — timescales of production
have liked to be in more contact myself with the engineering are out of line with timescales of development; low-cost
department about the project. At the same time, several voted sensing is not robust for the environment and application; data
for having a smaller overall team: Too many people have accuracy is questionable for low-cost devices due to factory
led to a lack of accountability and many misdirections in calibration (or lack thereof); and so on. We are not the first
communication. and The number of people involved on the ones to experience these: Peter Lyle and colleagues reported
project should stay a little more limited to ensure more on similar ones [
The potential communication improvement was best and semi-structured interviews in a community garden in the
summed up by this student: If we could start again, I would city of Brisbane in Australia. They confirm Odom’s [
There were suggestions for a different technical platform: technology should even be trying to facilitate these small comIf I had to start from scratch again I would definitely choose munal agricultural efforts, as put forth, for example: “Mate, the Raspberry Pi board over the Arduino Uno R3 because the Pi board functions as a computer with a built in module that would allow the user to log into a network as if the board was a laptop. Also the Pi is more recommended to use if your trying to complete a project that is as difficult as ours. Which required multiple components and communicating with other embedded systems. Furthermore, the garden setting could be improved: The plot would have done best in full sun as well as being deployed in the spring through summer. The placement of the plants and sprinkler could have been more organized and dispersed evenly for minimum amount of watering with maximum crop growth.
Student valued the project management experience (It was a good chance for me to learn more about how to lead a project and how to effectively deal with people and management.), the horticulture learning experience (I also learned more about how to take care of crops which can help me develop my own garden eventually.), and reported that they enjoyed being around like-minded people (It was nice to work with people who are devoted in lessening our water usage and improving water efficiency.).
One student pointed out the option of doing the experiment at home to save mileage: This kind of project would be better if the student was allowed to complete the project in one of the group members home. The reason is because it would allow the group to monitor the garden on a daily basis without having to drive to school especially if they live far away. There wouldn’t be any issues with the wifi regardless of what board the group uses because the user has complete control of their network.
All students gave us positive feedback about the experience,
for example The experience as a whole was very rewarding.
as well as Overall, it was fun visiting that part of the campus
and communicating with the various people involved with
this project. Finally, several students expressed gratitude, for
example I really enjoyed being apart of something and seeing
it grow. Working with Claudia and Julio was great. I feel
grateful Libby trusted in us and gave us the opportunity to
help run this project.
we don’t need a chip to tell us the soil’s dry”, by Odom [
Lastly, while we were enthusiastic about applying Mann’s Transformation Mindset Principles during the inception phase of the project and reflected on them later on, we do see that the implementation falls short in terms of answering the bigger questions that inspired Mann’s work. This could be mitigated in the future by developing an instrument that helps to tie the results of a project back into a debrief of the experience that includes a reflection of the Transformation Mindset Principles.
We used Mann et al.’s [
We are aiming for a replication in a Long Beach community garden, where we can interact with the general public and build further bridges between the university and parts of the local community that are not necessary likely to interact much with academia. While we would have wanted to replicate the experiment this year in the garden where the first run was conducted, the CSULB university garden will unfortunately be taken out in its current form for a new building and the future location is unclear at this point. We are currently in negotiation for the future plot.
In a related project, we would like to further the
development of a companion planting composer software [
In further future collaboration with hospitality management, we envision a cooking class supported by app development where cooks could capture their own recipes and have an educational platform helping them improve their skills and get feedback from each other. This collaboration faces the same time management challenges as the last one, as class times are prescribed by administration and instructors have little leeway for arranging additional mandatory meetings. Consequently, it will take a few strongly motivated and dedicated students to enable an exploration of this cross-departmental development and implementation.
In this paper, we reported on the set-up, discuss lessons learned from the pilot, and project future scenarios leveraging the sustainability transformation mindset principles that support transitioning towards teaching sustainable livelihood with the support of ICT. We harvested a significant amount of kale and lettuce, despite the fact that the automatedly watered boxes yielded less harvest than the manually watered ones.
While we consider the conducted action research successful
in terms of experience, insights and lessons learned for all
participants, we also take a critical look at the research. In
the discussion of “undesigning” [
Broadening to a wider perspective, we should also ask
“What if sustainability doesn’t work out?” [