Innovations in embedded computing are happening every day. We’ve collected stories from the leading developers who are making the most of NVIDIA Jetson and stretching the boundaries of what anyone thought might be possible. Maybe we’ll tell your story next!
Featured Partner: Prat Prem Sankar (NVIDIA Jetson Intern)
Prat Prem Sankar is one of NVIDIA’s youngest interns and spent last summer exploring deep learning and building humanoid robots with Jetson TX1. He wants to inspire more students to take on robotics and show that it’s possible for anyone to create their own robot.
Flight and robotics have always been Marc Gyongyosi’s biggest passions. Now he’s the founder of a startup, IFM, that’s developed a Jetson-powered autonomous drone. The company first unveiled its indoor navigation technology at NVIDIA’s GPU Technology Conference in April 2016, and is now staking a claim in the $5 billion warehouse analytics industry. Using Jetson technology makes his flying robots small, light and agile enough to zip through the corridors and tight spaces of a warehouse.
Saverio Murgia, Horus CEO and co-founder, was inspired to create his company two years ago after meeting a blind person on the street who asked for help finding a bus stop. The Horus wearable device uses computer vision and machine learning to aid visually impaired people through the help of Jetson. Now they can read a product label or a book, recognize a friend on the street and get help navigating street crossings and obstacles. Users can also customize the device so it recognizes objects specific to their needs. Learn more about this life-changing technology.
Aerialtronics is a manufacturer of technologically advanced drones for the commercial market that uses Jetson and deep learning software from Neurala. The company believes in a fully automated drone workflow for the safe inspection of cell towers, power lines, wind turbines and other infrastructure. Cell tower operators can dramatically increase safety, lower costs and collect more accurate data.
Stereolabs Brings Advanced Computer Vision Capabilities to 3D Mapping. In the drone industry, the weight and size of the main board is critical. With the ZED stereo camera by Stereolabs, developers can capture the world in 3D and map 3D models of indoor and outdoor scenes up to 20 meters. The small form factor of the Jetson TX1 enables Stereolabs to bring advanced computer vision capabilities to smaller and smaller systems. See what is possible when these two technologies come together in drones to power the latest virtual reality applications.
Powered by the Jetson platform, Percepto is working on computer vision solutions to enable autonomous flight capabilities in the new era of low-cost drones. The NVIDIA Jetson TK1 gives Percepto drones high computing power along with low power consumption all in a small form factor. "NVIDIA enabled us as a very small company to get the information and support that we needed to get a product based on this high-end computer in a very short amount of time," says Dor Abuhasira, CEO of Percepto. Check out the video to see what Percepto has planned for the future of autonomous flight in drones.
MIT professor Sertac Karaman is teaching his students about robotics, drones and autonomous vehicles using the Jetson platform. Karaman designed an attention-grabbing course in racecar building to show his students the power and potential of embedded computing applications. When the course is offered again in 2016, the students will be using the Jetson TX1 supercomputer to build robotic cars capable of racing at 20 - 30 miles per hour. Students will use the Jetson TX1 for its machine-learning inference capabilities to be able to detect other cars that are racing at the same time. If you happen to be in the tunnels at MIT, watch your step!
The Jetson TK1 has been an accelerant to the design and development process of the Jibo, the world's first consumer social robot for the home. The increased processing power of the Jetson TX1 will help the creators of Jibo to do some very special things with vision processing beyond simple facial recognition, such as the ability to leverage more sophisticated object detection leading to a faster and more responsive robot for consumers.
The NVIDIA Jetson TX1 supercomputer is designed to meet the demanding computational needs of visual computing applications like deep learning in the embedded space. Using the Jetson TX1, the Kespry Drone System is able to provide construction and heavy equipment companies with specialized asset tracking functions that ensure the logistics of each project site. In the example shown in this video, Kespry runs algorithms in real-time on their drone equipped with the Jetson TX1 to localize and classify vehicles, which improves the efficiency and safety of everyone involved.