This month, our team packed our bags and landed in Pennsylvania for the legendary PennApps—the nation's first student-run college hackathon.
Participating students, ranging from novices to seasoned hackathon professionals, engaged collaboratively in learning, competition, and the pursuit of honing their engineering skills while collaborating on exciting projects.
At this fall's hackathon, our game plan was clear: show the students what Viam can do, support them with their projects, and make the most of those intense 36 hours. So, here’s a peek into our journey.
Stepping onto the PennApps stage
New to Viam? Think of us as the software that makes your smart machines even smarter, from massive robots to your everyday gadgets. At PennApps, we were all in, helping students focus on the bigger picture—the impact their projects could make—rather than battling software-hardware connections.
To introduce the power of Viam to the attendees, we demoed “Tipsy: Arcade” which showcased the autonomous drink/snack carrier, aka Tipsy.
This smart machine is equipped with a camera to see and recognize people through a widely-used machine learning model, allowing it to interact smartly with its surroundings. And, using Python, Tipsy can head towards someone while dodging any hurdles with its ultrasonic sensors.
To highlight our app’s additional capabilities, we also trained Tipsy, using Viam’s ML Model Service, to dodge a Cyan Pacman Ghost (just us in T-shirts!) and to emit sounds as it retreated. It could also follow a pink ghost, changing sounds as it approached. Switching back and forth between the two models and scripts was a breeze!
Spotlighting the builds
As a sponsor of PennApps, we had a category “Best Use of Viam Software in Robotics” for submissions.
We wanted to see passion, teamwork, and, most importantly, a clear vision behind each project. How well did they pitch their idea? How well did they work together as a team? And, of course, how creatively did they use Viam?
Take a look at some of the winners and their submissions below.
Project LearnSafe: a smart detection machine focused on school safety
This project won the “Best Use of Viam Software in Robotics” award on their first hackathon attempt, totally nailing it!
Project LearnSafe is designed to enhance school safety proactively. It utilizes a smart machine trained on Viam to detect individuals carrying dangerous weapons in educational settings, immediately notifying emergency personnel of potential threats.
Check out some highlights of this team’s build process below or get more details on DevPost:
- Utilized Viam to configure their device, comprised of a Raspberry Pi and camera.
- Deployed a machine-learning model via Viam’s ML Model Service, trained to identify individuals holding dangerous weapons such as guns as potential threats. This model was trained on existing CCTV footage.
- Leveraged Viam’s Data Management Service to capture images of themselves holding a knife to train additional data directly within the app, again using the ML Model Service.
- Integrated Twilio to automatically call 911 when a person with a knife was detected. For testing, they called their own number to avoid accidentally contacting 911.
Each member of the team went home with a robot development kit featuring a fully equipped Viam Rover, 18650 batteries, an RPlidar and a 3D printed holder for it, a Raspberry Pi with SD card and a charger. But this project still isn’t over! The team hopes to further enhance their ML model, boosting the efficiency and effectiveness of its detection.
Aerovision Pro Max: facilitating the transition to self-driving in air traffic control
Snagging Viam’s Honorable Mention award is the project Aerovision Pro Max—because, let’s be honest, “Who wouldn’t want to move a rover with just a wave of a hand?”
Aerovision Pro Max brings ‘hands-free’ control to life, enabling a Viam Rover to move according to standard Aircraft Marshall signals—it moves forward, backward, spins right, spins left, and halts, all on a gesture! Ideal for those aircraft still relying on manual control while others soar ahead with self-driving tech, ensuring seamless transitions and ultimate safety during taxi operations.
Dive deep into the creation journey below or delve into more details on Devpost:
- Empowered a Viam Rover to interpret hand signals using Viam’s app, exposed APIs, and Python SDK.
- Tracked hand movements in real-time via webcam using an OpenCV model, translating each gesture to a specific rover move.
- Faced and overcame challenges, all to ensure the rover responds accurately to every hand signal.
Despite the seamless integration and success, the team is aiming higher—they’re gearing up to introduce more accurate pose tracking and a flurry of features for AeroVision Pro Max Deluxe. Keep waving, and stay tuned!
Eco Lens: a sustainable computer vision smart solution employing a camera for the recognition of recyclable objects
Rocketing into the top 20 of the entire hackathon is the innovative Eco Lens!
Eco Lens is your savvy environmental pal—it uses cloud-based machine learning to tell you whether that item in your hand is recyclable or not, gives you a bit of a rundown on it, and even keeps tabs on how many items you’ve consumed.
Dive into their green journey below or explore more of the project’s details on Devpost:
- Created a tech blend of hardware and software, connecting to the smart machine via Viam’s versatile software and adding a camera to it.
- Employed Java Spring and Javascript for a smooth back-end operation, allowing the seamless integration of Roboflow and Ultralytics and the creation of a user-friendly database.
- Designed a front end using HTML, CSS, and Javascript, ensuring a user-friendly interface brimming with helpful functions and buttons.
While the Eco Lens is already sprouting, the team sees it growing into coffee shops and cafeterias, post assessing its business demand. Eco-friendly and user-friendly, EcoLens is ready to leave its mark!
Mentoring teams to create impactful builds
Being a Developer Advocate brings me immeasurable joy, especially when I get to guide passionate teams through the intricate dance of melding software with hardware. Witnessing eight dynamic teams transform their visions into tangible smart machines using our platform, Viam, was truly amazing!
Viam turned the ‘Hows’ into achievable realities, offering an intuitive escape from the complexities of setups and allowing teams to dive deeper into actual development. It’s these journeys, these unfolding stories of passion and innovation, that are the true essence of my joy in this role.
All in all, myself and the rest of the Viam team present had a great time! Take a look at the rest of the submissions featuring Viam and get inspired by many of the projects built in just a short, fun weekend!
