The fondness towards gardening and plant growth was a major motivation behind our idea of a SICPI (Smart IoT Connected Plant Incubator). We wanted to build a feasible and versatile solution with a comparatively easier user interface that can be used by everyone and not only laboratories or industries.
Different species of plants have a different kind of requirement for growth. It’s often difficult to provide the right condition for a plant to grow in an open environment. By right conditions, we mean the amount of soil moisture, air humidity, temperature, light, soil pH, etc., that needs to be present in the environment with balance. With that being said, we have to monitor constantly an incubator. Unlike native ways, which required human labour to do so, we required to achieve the same more efficiently with IoT or RF, that also with fast response time.
Significance of the problem
In areas with extreme environmental conditions where it is not possible for many species of plants to grow, it becomes difficult to cultivate them. Also one of the biggest concerns, due to adverse effects of nature many plants are getting endangered. In the upcoming days, labs are going to test new species, which might require a whole lot of environment. So an artificial ecosystem is going to be something necessary for Science and Technology. Though many incubators are available in the market, most of these features either automatic lighting or watering or utmost both. Many of these lack autonomous control and monitoring, and even if such features are found, they are very costly, which constrains these incubators only to research and development works.
How the problem is being addressed today?
The problem we discussed is currently being addressed today due to the adverse effects of pollution. In spite of having good soil, plants cannot grow properly, and if grown, they don’t produce fruits properly. Many known scientific plants are getting endangered. The Plant Chambers that are available in the market are very expensive, thus constraining these chambers only for research and development purpose. These plant chambers have several places of improvement. The artificial ecosystem created by the Plant Chambers has to be manually regulated, which is a drawback as it requires the presence of an operator. Also, because of the high price of the Plant Chambers, they are not very locally popular. However, if these gaps are reduced, we can estimate a striking increase in the number of customers from hobbyist gardeners to professional nurseries to many more. Thus giving a wide variety of reach.
How are we planning to solve this problem?
We are planning to develop an autonomous Plant Incubator that will produce artificially controlled environments crucial for plant growth and seed germination in any environment. This solution will be useful for laboratory incubation of plants for research and to also keep endangered plant species alive. The Plant Incubator will precisely control autonomously environmental parameters, such as temperature, soil moisture, humidity, light source, CO2, and oxygen level. With IoT, we will be monitoring these parameters remotely. By means of connecting these with a database and running some algorithms, we can predict naturality. Due to the storage of many plant characteristics over time, we can study nature more effectively. We are planning to install cameras in our Plant Incubator that will allow us to monitor the plant growth remotely. The program will alarm the user of the plant getting not enough water or CO2. The program will also alarm the user about any diseased growth in the plant. According to the plant being grown inside the Plant Incubator, the user will be able to preset a few ideal conditions that will be maintained by the chamber starting from the seed plantation. The condition inside the Plant Incubator will be monitored using several sensors. The following are how we are going to provide the ideal conditions for plant growth:
- Fluorescent lights with large reflectors provided a relatively spatially uniform photosynthetically active radiation intensity.
- A portable air conditioner provided an ample cooling capacity, and a cooling water mister acted as a powerful humidifier.
- Temperature, relative humidity, and light cycle inside the chamber were controlled using IoT, which allowed the environmental parameters to be monitored and programmed through the internet.
- The environment parameters like the temperature, level of CO2, soil moisture etc can be programmed by the user and the Incubator will sustain these parameters autonomously, also it will keep it updated for monitoring.
Our proposed design (the tools we need to solve this problem)
For the hardware part - We are using an LM35 for temperature control, PT100 is also an alternative. For the humidity, we are using the AHT10 temperature and humidity sensor. Using more than one type of sensor, we can use Kalman filters to get accurate measurements. Using capacitive soil moisture sensor and soil pH sensor. With the MICS5524 air quality sensor, we will get the amount of CO2 and oxygen level inside the Incubator. With OPT3002 ambient light sensor, which offers sensing of a wide spectrum of incident radiations, will be used for configuring the light source optimally required for incubation. This was all about the sensing part. Now coming to the actuation part, for light sources, we will be using a different wavelength of light sources so that it can emit a wide range of wavelengths, from UV to Infrared. With the help of the ambient light sensors, we will be controlling the intensity of each lights to get the desired light ambience for optimal incubation. For the soil moisture, humidity, injection of CO2, solenoid valves will be used to actuate. For the temperature control, 80watt filament bulbs can be used for testing, but keeping in mind that the light from the bulb can interfere with the ambient light source, we will be using a heating element for keeping the incubator's temperature. And for the prototype, we will use a small radiator fan to cool the incubator. Coming to the controller part, we can use an STM32 microcontroller with analog multiplexers to take the readings from the sensors. However, most of the sensor's output and its reading will be in digital format, i.e. via i2c or SPI. To get the IoT connectivity, we can use a WiFi chip, or we can go for the popular ESP12F or ESP32 microcontroller that comes with an embedded WiFi. Due to the low resolution of ADCs and higher power consumption than STM32, we are perplexed.
For the software end- At first, we will keep it simple, and we can use google's spreadsheet to store the sensor readings. With the API credentials and key, we can fetch the data from the spreadsheet with an HTTPS request. We can make a desktop application using Java Swing that will be fetching the data from the spreadsheet and parsing the values, and showing it in a graphical order for further processing. Since the data will be stored in CSV files, we can use it for further ML models. The desktop application will have a set parameters area where the user can set the temperature and other environmental parameters required. The same will get updated to the sheet, and the microcontroller will get commanded. Since the incubation will be a slow process, we don't have to make a fuss about fast updating, and a 0.2Hz updating rate will do the job.
Is our solution sustainable? What is our strategy for making the solution sustainable?
Yes, SICPI is a sustainable solution since it requires not such complex operations in the field, which makes it robust and less prone to crashing. Thus making SICPI, as a product, very sustainable. Most of the heavy computation will be done at the servers or the cloud, which saves the architecture hierarchy. And yes, there exist Plant chambers and modified greenhouses, but they are quite expensive. The high cost of maintaining a controlled growth environment is often a limiting factor when determining experiment size and feasibility. To overcome the limitation of commercial growth chambers, we are planning to construct an inexpensive automated Plant Incubator with IoT connectivity that will not only monitor but can autonomously control the environment. Our business strategy involves approaching laboratories that need a modified Plant Chamber for incubation and complete observation of the plants for every second of the day. Organisations working towards saving several endangered plant species from going extinct will also be our target audience. The data collected by the Plant Incubator will also be available for the user to access any time, hence will allow the user to make a discovery about a plant.
Who will benefit from the solution proposed by us?
People like biologists, scientists who are conducting research on plants and the organisations working towards saving several endangered plant species from going extinct will highly benefit from our idea. Other than that, institutes trying to teach students about stages of plant growth and the ideal conditions required for different plants will also be able to use our idea. And last but not least, people like us, who have the hobby of gardening but don’t have the ideal environment for it, or the ones who don’t enough time to maintain the plants, will tremendously benefit from our proposed solution of the SICPI (Smart IoT Connected Plant Incubator).