TL;DR: The EVLA model enhances vision-language models for robotics by improving speed and reducing resource demands, Robotic Foundation Models will simplify the development of versatile robots, and the FCC’s new spectrum access expands drone capabilities.
The Edge Vision-Language-Action (EVLA) model focuses on improving the efficiency of vision-language models for robotics by enhancing inference speed and reducing computational demands, making it more accessible for real-time applications on edge devices like the Jetson Nano. The model streamlines processes by avoiding autoregressive constraints and leveraging smaller models without sacrificing performance. This advancement enables more practical deployment of intelligent robotic systems, allowing for faster, cost-effective solutions in robotics that can operate on less powerful hardware.
Major takeaway: EVLA democratizes advanced robotics by making efficient, real-time vision-language-action models accessible on affordable hardware.
Robotic Foundation Models (RFMs) will transform how robotics are built, bought, and funded by providing a new framework for innovation. Similar to how large language models revolutionized AI by making it more accessible, RFMs are doing the same for robotics. These models will equip robots with a generalized understanding of tasks like navigation and object manipulation. This shift allows robots to operate in a wider range of environments, reducing the need for task-specific programming.
Historically, robotics companies had to build full-stack solutions involving hardware, autonomy, and software integration. RFMs can simplify this process by offering a scalable and flexible base upon which developers can create application-specific robots without having to reinvent the wheel each time. This ability to generalize tasks and adapt to different scenarios is groundbreaking, moving the industry closer to creating versatile robots that can seamlessly switch between various roles and functions.
Major takeaway: Robotic Foundation Models will drastically lower the barriers for innovation in robotics, enabling more flexible, adaptable, and scalable robotic solutions across industries.
The FCC recently made a significant move by opening spectrum access for drones, allowing them to operate within the 5030-5091 MHz band. This decision is a game-changer for the UAS (uncrewed aircraft systems) industry, particularly for industries relying on drones for infrastructure inspections, logistics, and beyond-visual-line-of-sight (BVLOS) operations.
Initially, the FCC will enable limited, line-of-sight communications using this spectrum, crucial for establishing stable ground-to-air links for drone control. Over time, a more advanced, dynamic system will be rolled out to allow real-time spectrum assignment, modeled on successful systems like the Citizens Broadband Radio Service (CBRS). This phased approach aims to expand the use of drones in more complex missions, increasing operational efficiency and safety for commercial applications.
Major takeaway: This spectrum access opens up new opportunities for the drone industry, especially for more complex and large-scale operations, marking a pivotal moment for commercial drone integration.
About Lucid Bots:
Founded in 2018, Lucid Bots Inc. is an AI robotics company that is committed to uplifting humanity by building the world's most productive and responsible robots that can do dangerous and demanding tasks. Headquartered in Charlotte, the company engineers, manufactures, and supports its products domestically, which include the Sherpa, a cleaning drone, and the Lavo Bot, a pressure-washing robot. Lucid Bots' products are elevating safety and efficiency for a growing number of customers around the world. Lucid is a Y Combinator-backed company, with investments from Cubit Capital, Idea Fund Partners, Danu Ventures, and others. Lucid Bots raised $9M in Series A funding in May and was recently recognized as the 4th fastest growing manufacturing company in the United States.