Making the shift to more sustainable lifestyles and technologies is the most important it has ever been due to climate change, rapid loss of habitats, and plummeting biodiversity. This means everyone, especially those with higher carbon emissions, must do their part both individually and societally to solve these problems. Because September is Sustainability Month on Hackster.io, we have picked out some of the best environmentally focused projects that have been posted and cover everything from improving agriculture to keeping track of litter.
Fresh water is one of the most valuable resources we have on Earth, which makes waste a very large issue. Droughts across the Southwestern United States, Western Europe, and China have led to massive reductions in the amount of water available for irrigation, and Eric Hall's Precision Agriculture prototype hopes to be a potential solution for smaller farmers. His design is made up of an Arduino Uno-based sensor network that continuously monitors the current soil and weather conditions while simultaneously streaming that data to the Azure Cloud. The hope is to pinpoint where and when water is needed most so that overuse becomes a thing of the past.
Continuing with the theme of intelligent agriculture, Atul's Smart Greenhouse project aims to be a complete package when it comes to optimally growing plants. The greenhouse uses a combination of two Intel Edison development kits, environmental sensors, and an array of relays to precisely control the interior conditions. The first Edison sends sensor information to an AWS MQTT endpoint where it's then analyzed and made available for viewing on a web-based dashboard. From here, the second Edison can activate a ventilation fan, spray mist to increase the humidity, or change ambient light levels in the hope of improving crop yields without the need for carbon-intensive fertilizers.
Over the past couple of decades, there has been a noticeable decrease in the number of honey bee hives, most likely due to parasitic mites, pesticide, and warming temperatures. And in the United States specifically, bees make about one third of our food supply possible. To help improve beekeeping and hive management, Dave Veith came up with a live telemetry system that relies on a series of Raspberry Pi boards and digital temperature sensors. Temperatures are sent to an AWS IoT endpoint that, in turn, stores the data and generates alerts if the values fall outside of their normal ranges with the goal of saving any hives that get too warm.
As powerful storms and abnormal temperatures get more frequent, having the ability to produce high-resolution meteorological models has become a necessity. Evan Diewald took advantage of Helium's LoRaWAN-based wireless network in order to provide these detailed weather maps without the need for expensive hardware or software. But beyond merely providing a plan for assembling a Raspberry Pi 4 and Helium Developer Kit into a remote weather station, this system also rewards users for submitting their data. In essence, each node stores and submits its own weather data to the rest of the network via a blockchain and gets rewarded after the transaction has been verified. With this approach, weather readings can be made more accurate while also consuming less energy than a traditional blockchain approach.
In 2021, around 60% of the United State's electricity was generated using fossil fuels, making it a prime candidate for a green alternative. Nick Fielding achieved his goal of making a more efficient solar panel setup by incorporating just a few simple components into an automated sun tracker. He connected a couple of solar cells in series to a voltage booster that could provide the required 5V to a USB device, such as a cell phone. Below the panel is a servo motor and two photo resistors that face east and west, all of which are wired to an Arduino Uno. The board reads the voltages coming from the resistors and moves the panels accordingly so that they are always at the optimal angle for maximum power generation.
One of the easiest ways to save on heating and cooling costs is to prevent the thermal energy from being lost/entering the house in the first place. However, an open window can quickly lead to the thermostat switching on the heater or air conditioner for hours on-end. To prevent this from happening, Fike Rehman created Elara with the goal of alerting the homeowner if the window is open while the central HVAC system is active. It works by using a reed switch to check if the window is open, whereas a clever micro switch assembly detects airflow coming from a vent. Once the two conditions are met, an alarm LED activates and its local webserver can be queried for the current alarm status.
Forests around the world, namely the Amazon Rainforest, are experiencing unprecedented levels of illegal logging due to mining and farming operations, the consequences of which are devastating to the local ecosystem and the world at-large. Alejandro Sanchez has built a small device using the QuickFeather and Helium Development Kits that listens to ambient noises and can detect when logging is taking place with the help of machine learning. Once this activity has been recognized, it sends a message via LoraWAN to an AWS service for further analysis and display.
Composting leftover food and paper materials can be a great way to introduce nutrients back into the soil without the need for artificially produced fertilizer. Darian Johnson and Mike Bradford's smart compost system aims to make this process easier by using several sensors to log the conditions of the pile. Beyond merely showing the temperature, moisture, methane levels in a dashboard, the Intel Edison can even recommend certain actions for ensuring an optimal mix.
Maintaining the Earth's wealth of diverse mammalian species is important for the ecosystems they inhabit, but poaching threatens their existence. Chamal Ayesh's project combines an nRF52 Development Kit with a mobile app and Avnet's IoTConnect service in the hope of protecting elephants. Upon reading a value from its onboard accelerometer, the nRF52840 passes the information to an awaiting Android phone over Bluetooth Low-Energy. From here, the information can be stored and plotted in an IoTConnect dashboard.
This waste bin is several steps above the typical trashcan, as it is packed with sensors that can help track potentially catastrophic fires that could harm the surrounding air quality. Built by team KMUTT, the bin's lid contains a distance sensor for measuring how much garbage is being stored, temperature/humidity and flame sensors for detecting if a fire is likely or has already started, and a lid switch for collecting statistics on how frequently the bin is used. All of these modules were connected to an Arduino MKR Fox 1200, which can communicate the data wirelessly over the Sigfox system while also providing locations within 100 meters of accuracy.
Few things are more annoying to look at than piles of litter building up along roads, buildings, and sadly, hiking trails. Because cleanup almost solely relies on human volunteers to report their location, getting to the trash before it spreads further can be a large problem. However, Nathaniel Felleke hopes to improve the process with his litter heatmap project that utilizes machine learning and cloud technologies for increased efficiency. If trash is detected by an Edge Impulse-generated image recognition model running on a Raspberry Pi 4, it sends a cellular message containing its GPS coordinates using a Blues Wireless Notecard module. Finally, the locations of trash can be plotted on a map to see where it's piling up the most.
Air pollution, whether indoors or outside, poses a major health risk to everyone since it can increase fatigue and risk of chronic diseases. The BreatheRight device created by Sridhar Rajagopal aims to predict overall wellness thanks to its carbon monoxide, particulate matter, and environmental sensors. Beyond these, the AWS IoT EduKit also runs an Edge Impulse classifier that listens for any coughs or sneezes with its onboard microphone and takes it into account when calculating an overall wellness score.
When imagining a garden, most people will probably think of a small plot of land outside with rows of vegetables and fruits. But after becoming inspired by a TED Talk, Dmitirii Albot had the idea to design and build an automated plan growing system that could facilitate efficient cultivation from seed to maturity. The Grow It Yourself (GIY) houses LED grow lights, nutrient delivery systems, water sensors, and an air control mechanism, which work together under the control of an ESP8266 and ATmega32U4. Even better, the WiFi capabilities mean that users can interact with the GIY remotely.
Climate change has led to a rapid increase in the number of droughts across the world, thus making water restrictions a more common occurrence. According to the EPA, a running faucet can consume ten gallons of water in just five minutes, so Naveen came up with a solution based around the Infineon PSoC 62S2 Kit that can use machine learning to detect if a faucet is running. It starts by capturing audio from an onboard microphone and passes it to a SensiML machine learning model for classification. Finally, any instances of a running faucet are published to an MQTT topic that triggers a text message alert.
Beyond mapping where litter might be located, the LitterBug robot by Salma Mayorquin and Terry Rodriguez is able to clean up these small pieces of trash all on its own. It is based on a simple RC car chassis that has had its internals replaced by a more capable Raspberry Pi 3. A microphone, GPS receiver, accelerometer, and infrared camera help the LitterBug navigate around various obstacles in its environment and recognize where trash is so that it can be collected. The entire thing is even powered by renewable power with an added battery pack for night operations.
One of the biggest drawbacks to using solar panels is that they are high susceptible to even minor fluctuations in the amount of light they receive, meaning nighttime or clouds can have a major impact on their power output. The Cloud Motion Vector system, designed by HyperChiicken, aims to predict when the backup generator might be needed at a power plant in order to preemptively turn it on to meet demand. It incorporates light readings taken at various angles and a variety of algorithms running on a BeagleBone Black to send commands over Thingspeak for the backup generators, ensuring a constant stream of steady power delivery.
The final project in this list is a small, LTE-enabled module that can remotely monitor pipes for gas leaks and even fires. Nimish Telang used a Sony Spresense main board, an LTE extension board, and a camera module that runs machine learning inferences for fire/smoke event detection. In addition to the camera, a CO2 and flammable gas sensors were included to measure concentrations of harmful combustion byproducts.