Don't Tell Robots What to Do — Show Them
Artly is a robotics-software company. It’s created Artificial Intelligence (AI) to make robots easier and faster to train.
Traditionally, industrial robots need to be programmed for specific tasks. This process can require specialized engineers, considerable time, and significant expense. Artly takes a different approach.
Using technologies including imitation learning and Vision-Language-Action models, Artly enables robots to learn by watching humans perform a task. In some cases, a robot can learn a new skill in as little as 30 minutes, with no coding required.
For example, instead of writing software instructing a robot exactly how to pick up an object, pour a drink, or manipulate a tool, a worker could demonstrate the task. Artly's software captures the movements and uses them to train the robot.
The potential advantage is flexibility. A robot that can quickly learn new skills could be deployed for many more jobs without requiring companies to build new automation systems from scratch.
Artly is already using this approach to teach robots skills including slicing ingredients, peeling fruit, garnishing drinks, and performing tasks requiring two hands.
The company has also introduced “Robot School,” a training platform aimed at helping robot manufacturers adapt AI models to their particular hardware. Artly says the system combines human demonstrations, simulation, reinforcement learning, and real-world operating data to train robots for complex manipulation tasks.
This technology has already been tested extensively in the real world.
According to Artly, its robots have generated millions of dollars in revenue and served more than one million drinks. The company has deployed its technology in dozens of real-world locations, and its robots have operated with organizations including Nvidia, Microsoft, Salesforce, Mayo Clinic, Intel, Cisco, Tesla, and MUJI.
That experience could be important. Many robotics companies are still developing prototypes in controlled environments. Artly's robots have spent years operating around actual customers, providing the company with a growing supply of real-world training data.
Founded in 2021, the company initially made its name with the Barista Bot, an AI-powered robotic system capable of preparing coffee drinks and even pouring latte art. But Artly’s ambitions now extend far beyond coffee. Artly is now looking to take what it learned from Barista Bot and apply it to a much larger market.
Its technology could potentially be used in manufacturing, logistics, food service, retail, and healthcare — industries where companies are increasingly exploring automation to reduce costs and address labor shortages. Artly isn't trying to manufacture every robot itself. Instead, its longer-term opportunity is to provide the intelligence and training technology that enables robots from different manufacturers to learn useful skills.
The company is led by Founder and CEO Meng Wang, a longtime artificial-intelligence entrepreneur. Prior to Artly, Wang founded computer-vision startup Orbeus, which was acquired by Amazon. He subsequently worked at Amazon Web Services and Amazon Robotics, experience that helped lay the groundwork for Artly's technology.
Artly has also attracted backing from venture investors including SV Tech Ventures, Amino Capital, LDV Partners, and Amitop Ventures.
Meng is a machine-learning expert with a background in AI and computer vision.
He began his career at the Mitsubishi Electric Research Lab, designing and developing new image and video resizing methods. From there, he joined Google as a software engineer, where he worked on imaging and video technology for the company’s Search feature.
Meng then co-founded Orbeus, a visual-recognition startup that was acquired by Amazon. Following the acquisition, Meng spent seven years with Amazon Web Services as a principal scientist before leaving to focus on Artly full-time.
He earned a Bachelor’s degree in Electrical Engineering from Shanghai Jiao Tong University in China, a Ph.D. in Electrical and Computer Engineering from Boston University, and he also studied at Stanford.
Yushan is an experienced product manager and technical leader. She has a background in robotics and AI.
After earning her Bachelor’s degree in Computer Systems Networking from Beijing University of Posts and Telecommunications, she worked toward her Ph.D. in Robotics at Boston University, which she obtained in 2013.
From there, she worked with Meng to launch Orbeus, the visual-recognition startup acquired by Amazon. After the acquisition, she stayed at Amazon for four years as a senior-product manager, leading initiatives including the company’s Alexa voice service and AWS Rekognition, a deep-learning computer vision service for image and video analysis.
Alex has experience scaling businesses. Notably, he advised Chinese video-game giant Tencent
on its expansion plans for WeChat, a social and messaging app that at one time was the world’s largest standalone mobile app with more than one billion monthly users.
In addition to his role with Artly, he is Founder and CEO of RSYNQ Corporation, a company building software for AI models and sensor-enabled devices. Its software enables robots and smart devices to securely connect with autonomous agents, allowing companies to deploy physical AI at scale.
Previously, he co-founded Realhand, a Palo Alto-based robotics company focused on creating human-level hand functionality for robots.
Alex holds a Bachelor’s degree and Master’s degree from UCLA.
A VC firm with more than $1 billion under management.
A Palo-Alto-based VC firm investing in early-stage startups.
Just enter your e-mail address below & get this guide instantly
No thanks, I’d like to continue making next to nothing in the stock market.
