Congratulations to the Winners of Infineon’s Build AI for the IoT Contest! –

Last month, winners were announced for the Build AI for the IoT contest with Infineon. This contest had over 300 participants from 51 countries, 100 hardware applications, and 25 fully documented, qualified final project entries. The Infineon judging panel commented on how they were blown away by how different projects implemented machine learning elements and impressed by how much time and effort went into building and documenting in each project — content that will be extremely useful for other community members to learn from and informing the Infineon teams on what types of technology best resonates with developers.

Let’s take a look at some of the winning projects.

1st Place: Illegal Logging and Fire Detector

This project by EdOliver, Victor Altamirano, and Alejandro Sanchez utilizes the PSoC™ 6 62S2 Wi-Fi BT Pioneer Kit to recognize the sounds of falling trees, chainsaws, and human voices in protected environmental areas to detect illegal logging. The authors described how in their home country of Mexico, 70% of wood consumed comes from illegal logging activities, making this project a relevant solution in the authors’ community and a great example of how the open-source developer community finds ways to overcome challenges to our environment.

The PSoC™ 6 62S2 Wi-Fi BT Pioneer Kit is used to obtain audio signals which pass through a neural network via a SensiML model. This informs users of what kind of noise was detected, or if smoke from a fire was detected from the MQ135 sensor. A web app then shows on a map where the noise or smoke was detected, giving enough information to environmental conservationists to react quickly to the situation. The machine learning algorithms implemented on SensiML combined with the use of AWS as a cloud service allow this system to be energy efficient, therefore a viable, easily-reproducible solution for any area where illegal logging is a problem.

2nd Place: PSoC™ 6: Voice Controlled Stove with Edge AI

Shahariar created an edge AI smart stove that can be operated by voice command and interactive voice feedback without needing to be connected to the internet. This project is a convenient way to easily adjust a stove if a person is busy cooking several things at once and have their hands full. This project uses a voice recognition model and the PSoC 6 MCU to control stoves turning on and off, temperature adjustments, and timers for individual burners. While voice-initiated assistants such as Alexa and Google Home could be programmed with hardware to do something similar, they require a constant internet connection and force a user to download a mobile app – not ideal for those sensitive to data privacy.

3rd Place: Running Faucet Detector and Alert System

Naveen designed a smart device that uses a machine learning model to detect a running faucet, initiating an SMS alert using the ultra-low-power PSoC 6 MCU. For people wanting to become more conscientious of their water use, this is a great way to remind you and your housemates to be aware when they let faucets run. According to the EPA, leaving a faucet on for five minutes can waste up to 10 gallons of water, so this project is a sure way to help people become more aware of situations that would waste water. The IoT Sense Expansion Kit’s microphone is used in this project for audio sampling and an OLED display to show inference output. If a faucet is left on, the device will detect the audio of the water running and alert users so they can quickly react, turn off the faucet, and conserve water.

Winner celebration at Sensors Converge

The Infineon team announced the winners of the Build AI for the IoT Contest at the 2022 Sensors Converge conference in San Jose, CA on June 28th. Hackster’s Alex Glow joined the celebration of the winners and had the chance to meet Gary Sugita, Director of Solutions for the IoT Compute and Wireless business unit in the Connected Secure Systems division of Infineon Technologies. He leads efforts to create hardware and software systems and reference designs that demonstrate IoT use cases with Infineon technology Mr. Sugita was one of the judges in this year’s contest.

What’s next?

You can learn more about how Infineon’s featured contest hardware by checking out the getting started resources: PSoC™ 6 62S2 Wi-Fi BT Pioneer Kit and the IoT Sense Expansion Kit. Learn more about ModusToolbox™ Software on the official page and browse through their lists of trainings, webinars and documentation to become a machine learning expert.

Thank you so much to everyone who registered, applied for hardware, and submitted projects in the Build AI for the IoT contest! Also, a big thank you to Infineon for providing cutting-edge hardware that so many participants used to create innovative and valuable projects with a wide variety of applications both inside and outside our homes.

Follow Infineon’s platform hub to stay updated on their next Hackster design challenge and upcoming developer activities around ModusToolbox and other featured hardware.

Spread the love

Leave a Reply

Your email address will not be published.