好色先生TV

FieldAI builds field-proven embodied AI software 鈥� an 鈥渁utonomy brain鈥� that helps mobile robots operate reliably in the real world. Our Field Foundation Models (FFMs) are embodiment-agnostic, allowing the same core intelligence to scale across various robot types and missions. We enable mapless, GPS-denied autonomy, allowing robots to navigate and make safer decisions in dynamic, unstructured environments without pre-programmed paths. Our technology is built to take robots beyond controlled demos into real-world operations at scale.

CMU robotics has looked to the stars for decades. This model is a full-scale replica of , the university鈥檚 next venture into space. Slated for a 2029 launch, the rover will advance our understanding of water on the moon as it autonomously explores near the lunar south pole.

Pittsburgh is no stranger to murky waters, and CMU research enables robots to generate detailed models when visibility fails. In this demo, an underwater robot performs live mapping of a submerged structure using sonar. This technology could improve inspection of bridges, dams and critical infrastructure.

Tracking invasive species across large regions is time consuming and costly. CMU researchers designed a shoebox-sized mapping payload that can identify trees of heaven in the wild. The team has mapped the invasive species on bike paths along Pittsburgh rivers, including those near the RIC.

Step inside the game. This demo lets fans watch live sports from different vantage points, like sitting courtside, hovering overhead or standing right behind the players. Real-time rendering lets viewers instantly replay highlights from any angle, creating an immersive live stream that brings the energy and atmosphere of the stadium to fans at home.

This research involves creating a smart drone that can fly autonomously, without GPS, to avoid obstacles. We hear about drones everyday, but many of these drones lack the ability to avoid obstacles in an internal environment like a tunnel. We accomplish this by giving our drone an eye and a brain. The drone uses a camera to create a three-dimensional map of the environment and determine its location. Our algorithm predicts the movement of objects and effectively avoids collisions.

This team has developed robotic and computing systems that mimic the rich versatility and robust mechanical properties of natural organisms. These include bio-inspired robots capable of various novel methods of locomotion as well as grippers that mimic the ability of an octopus tentacle or human hand to grasp and manipulate objects. Such efforts also include new classes of wearable electronics and physiological sensing technologies for human-machine interaction and health care monitoring.