news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

recently, the embodied artificial intelligence company "swiss-mile" announced the completion of a $22 million seed round of financing. it is worth noting thatthis round of financing was jointly led by sequoia capital and jeff bezos through bezos expeditions, with participation from amazon industrial innovation fund, armada investment and existing investor linear capital.

swiss-mile was founded in april 2023 and focuses on the research and development of embodied intelligent robots. the company hopes to create the interactivity and generalization of robots, make them universal, and help people get rid of trivial housework and daily work.

swiss-mile is backed by the robotic systems laboratory of the swiss federal institute of technology in zurich and is a subsidiary of the school. before founding swiss-mile, founder marko bjelonic systematically conducted research on robot intelligence and wheel-legged robots.

as a doctoral student at eth zurich, marco had previously studied systematically with professors marco hutter and stelian coros of the eth, as well as professor sangbae kim of the massachusetts institute of technology (mit).

marco mainly studies the design, mobility, autonomy and controllability of robots, and integrates the laboratory's research results related to artificial neural networks into quadrupedal and wheeled robots to enable robots to better realize embodied intelligence.

currently, the quadruped robots developed by swiss-mile can stand and walk on two feet, and the wheeled robots can drive automatically and grab packages.

"swiss-mile" autonomously trains robots to adapt to and judge the environment, and enabling robots to perceive the environment is also the first step for them to eventually enter the family as nannies.

the company integrates reinforcement learning and supervised learning into a unified ai training framework, allowing robots to autonomously learn environmental data from scratch and continuously adjust their learning progress based on the specific environment.

this learning method allows the robot to continuously learn new knowledge from the environment and continuously iterate the way it serves people, gradually training the robot from a housework novice to a housework expert.

the robot-independent multifunctional ai framework can make all robots equipped with ai brains smart and obedient. in just one week, swiss-mile successfully adapted the ai ​​brain developed by the company to different robot hardware, giving them a new neural network.

the device can last about 5 hours after charging and has a payload capacity of about 60 kilograms.

currently, the application scenarios of "swiss-mile" are mainly concentrated in the fields of monitoring and logistics, helping couriers solve the last-mile delivery problem.

investor comments:

caroline fu, partner at sequoia capital, said: "swiss-mile has deep accumulation and industry knowledge in cutting-edge algorithms and comprehensive solutions. we are pleased to see the company's technological progress and believe it will play a leading role in the intelligent industry transformation." sequoia capital invested through a us dollar fund composed of multi-national lps.

harry wang, founder and ceo of linear capital, said: "swiss-mile is one of the world's leading embodied ai companies, founded by pioneers of this trend. what excites me most about swiss-mile is that they have an international and customer-centric mindset from the beginning. they will be able to leverage their global resources, making swiss-mile a leading force in the next wave of the physical ai revolution."