A Community For Dads.
Planting Stems //
Planting Stems uses physical computing and data visualization to teach students about the greenhouse effect.
A low cost, hands-on science lesson for middle schoolers
10 weeks (January 2019 - March 2019)
Usability testing, interface design, 2D fabrication,
lesson plan design
Prototyping Studio, MHCI+D Program
Samantha Baker, Beijia Wang
Arduino, InDesign, Processing, Illustrator,
Laser Cutter, Premiere Pro
Microsoft Education's Hacking STEM projects aim to build affordable inquiry and project-based activities to visualize data across the STEM curriculum. We were tasked with developing a new lesson plan that:
Cost under $10 a student (excluding cost of Arduino Unos)
Met Next Generation Science Standards (NGSS)
Incorporated physical computing and data visualization
Planting Stems uses Arduinos + temperature and humidity sensors to allow kids to monitor plant growth over time in three different environments: open air, greenhouse, and "climate change" (a greenhouse with added CO2).
A hands-on way to build scientific instruments
Arduinos take the "black box" out of typical scientific instruments and become a learning opportunity in and of itself.
Built with teachers and students
Each detail and step of the lesson plan was designed and tested with real teachers and students.
Uses cheap, everyday materials.
By using materials that one can find in any grocery store, the cost for each student is just $8.75 per student.
Ideation & Research
We set off to learn more about NGSS. Since it had been a while since we were all in middle school, we also took some time to look at what a "normal" lesson plan looked like in terms of timing, difficulty level, concepts covered, etc. As a group, we generated ten lesson plan ideas that met certain NGSS standards. After reviewing the lesson plans that Microsoft already had, which were largely physics based, we were inspired to tackle a biology lesson and landed on building a lesson plan around teaching students about the greenhouse effect.
Our basic idea was to have students grow plants under various conditions to watch the greenhouse effect happen in real time, and to better understand how conditions like heat and humidity affect a plant's health.
Since our two most important stakeholders for this program would be teachers and students, we set out to prototype and test with both groups.
After outlining a rough lesson plan idea, we did our first prototyping session with University of Washington's KidsTeam, a participatory design initiative in the Information School's Digital Youth Lab. We went into the session with a few questions and goals:
1) Gauge the general background knowledge level of the students.
- What have they learned about plants in school?
- Are they familiar with the general processes of the life cycle of plants?
2) Is the experiment interesting to them? Are they excited?
3) Is our method of teaching the core concepts engaging?
We had the students run through the process of "planting" their seeds in the plastic bags as we asked them various questions.
From this we learned that the kids, who were even younger than our target audience, actually knew a lot about plants already. They could tell us that humans breathe air and plants "breathe" carbon dioxide, and some of them had even grown plants before. That gave us confidence that we could up the complexity of the experiment and kids could likely follow. The students were also really excited about growing something but impatient about seeing progress. We left the session with the following questions, and then set out to do further testing. Going into our sessions with teachers we focused on the following questions:
Is there a way to speed up the process to make the experiment more engaging?
Will students be able to make the conceptual jump from the physical world to the data they will see?
How do we make sure the lesson plan is accessible to all students with varying degrees of background knowledge?
We knew that no science experiment would work if it didn't work for teachers. We were interested in hearing about how the realities of the classroom may come into play with our concept.
We spent a week mocking up various low-fidelity data visualization ideas together. To speed up the process we made them in Sketch and then simply simulated the graph movements by flipping through the static screens on our laptop.
My teammate finalized the Arduino code to get our sensors up and running and we then spent the following two weeks conducting 5 prototyping sessions with teachers with our low-fidelity data visualizations and a rough draft of our lesson plan, each taking turns being lead facilitator.
We learned a lot from the teachers, especially about catering to different student needs. For instance, because science isn't testing on a state or national level, some elementary schools are forced to cut science classes all together, which leads to extremely different science backgrounds for students entering middle school.
Their feedback was invaluable in shaping our subsequent versions of both the lesson plan and the data visualization.
Added in flexible parts of the lesson that would give teachers the ability to increase or decrease the difficulty.
We reframed the lesson plan to make it about the student's solving a real world issue with the data they collect.
Added a potentiometer into our Ardiuno set up so students could plot their plant growth by turning the potentiometer instead of just manually plugging in the data.
With this feedback in mind, we were able to craft our final lesson plan and move our Illustrator prototype into Processing.
Students Round 2
Our final KidsTeam session presented us the opportunity to do a full test of our instruction booklet. Due to some complexities in how our Arduino sensors were coming together, our final schematic ended up looking pretty complicated and we were a bit worried about student's successfully wiring everything, and knew student's would likely need to be able to do this successfully in under 30 minutes for a typical class to stay on schedule.
We watched the students follow our instruction booklet and build the Arduino circuits. The students were faster and more proficient than we had guessed and it was exciting to see them so engaged in the activity.
We saw late in the process that they had mixed up the Power and Ground lines, a common mistake even for adults, and had to correct them.
In our final version, we rewired the entire breadboard and put the two lines on opposite sides of the board to ensure that the mistake was less likely to occur since it would be difficult for a teacher to diagnose what had happened if this occurred and the student's sensors subsequently got fried in the process.
For the next version we also made the diagrams much bigger so multiple students could view it at once.
Designing the Interface
The big pull of this project is that students were crafting their own scientific instruments that would enable them to collect data. An important part of that was how the data coming from the sensors would show up on screen.
I designed our screen in InDesign. From there it was as simple and finding the screen coordinates and making out Arduino output map to a rectangle that would grow proportionally with the data, creating a graph.
Final Lab Demo
For our presentations at Microsoft, each group was asked to craft a high fidelity lab demo version of their experiment. We elected to laser cut clear acrylic to form unique plant containers in place of plastic bags, as well as an Arduino enclosure.
Don't Underestimate, Go to the Source
We almost wrote off our idea as "too complicated". The schematic of our Arduino diagram looked complex even to me, and the director of KidsTeam told us that we shouldn't get our hopes up. But we tested it anyway and to our surprise they built the circuit about as fast as we had. To watch your assumption get blown up in your face is a great feeling, a validation of why we research and prototype.