I kill plants
I drown them. I bake them. I overfeed them. I starve them. Of sunlight. Of water. Of nutrients.
They stay strong for so long, but ultimately they shrivel up and die. Or their stems rot. Or they are eaten by bugs.
I never meant to kill them. I’ve always intended that they thrive and bear fruit. Still, death happens and I never understood why. Like so many serial plant killers, I just bought new ones.
But I liked my plants and never wished ill upon them. After all, they provided a ton of value to me: kept my air clean, gave me a delicious snack, and made the room brighter. My life is better with my houseplants.
I couldn’t help but feel they deserved better.
I needed to figure out why they kept failing: after all, my nearly 100% track record of killing plants indicated a systemic failing, not a corner case. I could have spent hours pouring over plant books and researching each species’ needs and wants. I could have thrown out the plants and bought fake ones. I could have hired a gardener. My issue, though, was that I really didn’t understand the root cause of the problem: any solution I came up with would have really only been trial and error.
What was the problem I was really trying to solve?
So often we’re faced with an ambiguous problem without an understanding of where to start. During these times, I remember a trusty friend who has systematically helped me solve many an issue: DMAIC. DMAIC is a data-driven improvement cycle used for improving, optimizing and stabilizing processes. There are five interconnected phases:
- Define. Clearly articulate the problem, goals, project scope, and customer requirements.
- Measure. Collect relevant data on the current process performance and identify key metrics for improvement.
- Analyze. Thoroughly examine the collected data to identify root causes of problems and areas for improvement.
- Improve. Develop and implement solutions to address the root causes and improve process performance.
- Control. Establish monitoring and control mechanisms to ensure that the improvements are sustained over time.
Perhaps it’s overkill to apply an analyze>plan>do style process improvement approach developed for reducing manufacturing errors to 3.4 defects per million opportunities to saving my defenseless little plants from the hands of a killer…
… but this simple example demonstrates how broadly applicable DMAIC can be. Its power lies not in solving the problem, but in helping you frame the problem to solve in terms of the data that can help you solve it.
So, on to the application …
Define
The DEFINE step of DMAIC is the initial phase where the problem, goals, scope, and key requirements are clearly articulated to establish the project’s purpose.
In the case of my houseplants, I identified that my plants were not thriving (100% mortality rate!) My goal therefore was straightforward: to improve the overall health and growth of these plants.
My efforts specifically focused on my indoor potted plants, including notable examples like my coffee tree and plumeria. I noted that plant health is a qualitative output, and I needed some quantitative key controllable inputs (aka Key Process Input Variables).
Controllable inputs are values that I could tweak as I measured, analyzed, and improved.
My inputs, therefore, became light exposure, watering amount & frequency, nutrients, humidity levels, and ambient temperature.
Defining these parameters provided a foundation for systematically saving my plants through targeted solutions in subsequent DMAIC steps.
Measure
The MEASURE step in DMAIC involves collecting accurate data, establishing benchmarks, and identifying key metrics to quantify the current state and better understand the problem.
For my houseplants, I actually needed a way to collect data about my inputs to pair with my qualitative output observations of plant health (i.e. wilting and browning leaves). Some quick Amazon research revealed a soil sensor to continuously capture critical environmental parameters: soil moisture, light intensity, humidity, and nutrient levels. I connected the sensor to my Home Assistant so I could measure these values in near-realtime.
Capturing these metrics provided an objective baseline to evaluate my plants’ conditions in the next DMAIC step.
Analyze
The ANALYZE step in DMAIC involves evaluating collected data to identify underlying causes of problems.
I analyzed the environmental data collected from sensors and compared them against optimal care guidelines obtained from a trusted crowdsourced database, Open Plantbook.
By charting my key inputs—soil moisture, lighting, nutrients, temperature, and humidity—over time against the recommended minimum and maximum thresholds, I could quickly visualize discrepancies and isolate potential root causes for declining plant health.
Through this simplified yet effective approach (adhering to the KISS principle), I identified three glaring issues:
- inconsistent watering practices,
- insufficient light exposure, and
- inadequate nutrient levels.
Further, I was able to rule out disease, humidity, and temperature related issues.
Although advanced statistical tools like control charts, Pareto analysis, or correlation tests could provide deeper insights for more complex scenarios, in this case, straightforward visualization clearly pinpointed the likely root causes impacting my plants.
Improve
The IMPROVE step in DMAIC involves generating and implementing targeted solutions to address identified root causes.
With clarity about the underlying problems affecting my houseplants, I developed and immediately started applying specific solutions:
- Watering. I adjusted the watering frequency and quantity according to real-time soil moisture readings.
- Light exposure. I moved my plumerias to brighter locations to optimize their exposure to natural light, while supplementing my coffee plants’ illumination with new grow lights.
- Nutrients. I acquired and applied time-release fertilizers tailored to each plant’s needs.
Throughout this phase, I consistently monitored relevant input metrics to verify that these corrective actions effectively aligned environmental conditions with the established care guidelines for each plant species.
I observed any changes in output as well, and happily noted that my plants looked greener, stronger, and overall more vibrant.
Control
The CONTROL step in DMAIC involves establishing systems and processes to sustain the improvements made.
Now that my plants were actually healthy, I refined my dashboard in Home Assistant so I could quickly monitor my plant’s metrics for each plant from anywhere.
The system provides immediate alerts whenever conditions drift outside defined acceptable ranges. I still need to manually apply corrective adjustments to lighting, watering, and humidity as needed–automation is possible, but I rather enjoy caring for my plants by hand.
Additionally, I wrote a quick 1-pager documenting some standard operating procedures covering watering schedules, fertilization plans, and pest control methods, ensuring consistency and repeatability in my plant care routines.
As I learn more about each plant, I can revisit the doc and tweak the inputs and process.
Conclusion
By applying the DMAIC framework, I was able to systematically diagnose and address the issues that were causing my plants to perish. Through defining the problem, measuring key variables, analyzing the data, implementing targeted improvements, and establishing control mechanisms, I transformed myself from a serial plant killer into a successful plant parent.
My experience served as a reminder that even seemingly simple problems can benefit from a structured, data-driven approach, and that DMAIC can be a valuable tool for improving any process, whether in a manufacturing setting or in the comfort of your own home.
Remember, the key to success lies not just in finding solutions, but in understanding the root cause of the problem and using data to guide your actions.
I imagine my plants appreciate the effort … I know I certainly appreciate them!