From robots to CNC machines, we offer many types of motion control solutions.
Platform for programming any industrial robot
Tool for modeling simple objects and complex environments
Bin-picking system with real-time motion planning
Laser cutter powered by motion planning
H 452mm x W 172mm x D 645mm
Power consumption: 300W
Cores: 12 core (24 threads)
Core clockspeed: 2.00Ghz
HDD Capacity: 500Gb
Motion planning for robot requires avoiding collisions with obstacles while optimizing the robot task time. Instead of doing the traditional labor-intensive human teaching, Mujin Controller Sim automatically plans efficient, fast, and collisionfree trajectories and generates robot programs that are 95%+ accurate for execution on the real robot. With motion planning, previously impossible tasks such as factory layout optimization and multirobot motion planning become possible.
Mujin Scanner is the ideal choice for doing a quick and simple 3D scan of an object or environment. A handheld 3D sensor developed specifically for robot simulation applications is used to capture 3D data. Scanning takes only a minute and the generated 3D model can be imported into simulation software immediately.
Mujin consists of experts who can apply the state-of-the-art technologies to design the most suitable robot system for industrial applications. In additional to robot motion planning, we have expertise in computer vision, software, electrical, and mechanical system design.
Mujin is able to utilize the full potential of existing hardware with our software. If the robot hardware is capable of moving faster and perform more complex motion, but is limited by its controller, we offer help to design a better controller that can realize the hardware's full potential.
Mujin has a lot of experience working with complex robot systems such as dual arm robots and 6-DOF robots on linear stages. We offer customers precise simulation that describes how real robots will perform the desired task and how the work cell should be organized.
Mujin appreciates and respects existing systems working on the factory floor. We are capable and more than happy to salvage and upgrade existing legacy systems.
Mujin develops technology to generate extremely robust and fast programs to solve analytic robot inverse kinematics equations. Our framework supports any robot structure, even when it involves closed links. Thanks to the flexibility of our inverse kinematics framework, we are able to plan motions for more than 100 different robots. We offer motion planning service for any robot.
Mujin has applied advanced motion planning technology with fluid simulation to quickly and safely generate trajectories for handing fluids without spilling them. We demonstrated this technology at IREX back in 2013.
Robot programming today is inefficient and makes up at least 40% of total system integration cost. To program a robot, the user has to manually move the robot joint by joint and record the trajectory for playback. The process is time consuming and there is no guarantee that the motion is optimal. It is difficult to optimize factory layout, as every new configuration requires a new teaching process.
Motion planning for robot requires avoiding collisions with obstacles while optimizing the robot task time. Instead of doing the traditional laborintensive human teaching, our system automatically plans efficient, fast, and collisionfree trajectories and generates robot programs that are 95%+ accurate for execution on real robot (existing simulators have less than 60% accuracy). With motion planning, previously impossible tasks such as factory layout optimization and multirobot motion planning become possible.
Inverse kinematics (IK) is a key operation in motion planning. Given the desired goal position of a manipulator, IK solves for the joint angles of the robot. Mujin develops technology to generate extremely robust and fast programs to solve analytic robot inverse kinematics equations. Our framework supports any robot structure, even when it involves closed links. Thanks to the flexibility of our inverse kinematics framework, we are able to plan motions for more than 1000 different robots.
Optimizing the robot placement and other objects in the workspace can quickly cut the task time (takt time) by 25% or more. However, layout optimizations are computationally intensive. Mujin utilizes parallel computing infrastructure to shrink the computation time drastically. For example, a layout optimization that takes days on a single computer can finish in minutes in the cloud.
Mujin generates low level robot programs from a high level industrial task language (ITL). We design ITL to be able to model any industrial task, so that users can focus on designing the actual manufacturing process. We believe that ITL is the key to unify all robot languages and make robots easier to use.
3D CAD models are used to model the world for motion planning. Among many existing CAD formats, Collada has recently become an ISO standard. Mujin maintains the robot extension to Collada and expects it to become the international standard for industrial robotics.
Mujin provides a RESTful web API for its core motion planning service. Outside systems can easily integrate with Mujin’s motion planning service using any platform in any language. An intuitive user interface can be built easily leveraging the power of a modern web browser and Mujin’s web API.
If you have any questions on our technology or available services, please contact us!
Issei is determined to leverage Mujin's automation technology to fundamentally change the landscape of manufacturing. Before founding Mujin, Issei was an award-winning sales manager at the Japanese branch of Iscar, one of the most profitable precision tool makers in the world. The years spent running on the factory floor in Japan, Korea, and China gave Issei deep understanding and appreciation of manufacturing automation, and inspired him to take on this billion dollar challenge.
Rosen's one and only passion is to increase worldwide quality of life by deploying as many automation systems as possible into the real world. After getting a PhD from the Robotics Institute at CMU, his dreams led to co-founding Mujin, a robotics startup employing advanced AI technologies, in the world's most hardcore and intense manufacturing environment: Japan. He is a true believer of open-source and has founded and developed the widely used robot motion planning system OpenRAVE, which comprises the core of the Mujin technology stack and is used by many robotics labs around the world.
Chief of Software Development
Huan has a life long passion in robotics. He grew up in Shanghai, China, where electronic components were very accessible and cheap. Huan built many robots with his friends and participated in every local and national robotics competition that was available. Later at MIT, he majored in Computer Science and took part in Artificial Intelligence researches. After graduation, he learned a great deal about software development in business settings at Oracle, attended a hardware incubator in Shenzhen, and is now chasing his dream at Mujin.
A tinkerer at heart, Luke has been building robots and tearing apart machines since the 4th grade. During his time as an undergraduate, Luke developed an experimental X-ray optic production facility at the Havard-Smithsonian Center for Astrophysics as well as an intra-wing robotic inspection system for Boeing. At Mujin, he is the resident web developer and responsible for the user interface, API design and infrastructure for Mujin’s revolutionary industrial robotics platform.
The first of Mujin's many interns, Shohei has been applying his knowledge of mechanical engineering and computer science to prototype and build applications using Mujin's technology. After receiving his Bachelor's degree at The University of Tokyo, Shohei is currently working on his Master's thesis at Mujin.
Although he started his carrer as pipeline engineer in Nippon Oil Corporation, after his hard study, he succeeded to get qualification of certified accountant. He experienced a lot of audit to help companies to go IPO. Since he strong passion to create something from zero, he joined MUJIN to build solid management/administrative system by using his professional knowledge.
Computer Vision Engineer
Jeronimo envisages a world where robots seamlessly integrate with human life, by accurately perceiving their surroundings and autonomously responding to human needs. While pursuing his PhD in Computer Vision and Machine Learning for Robotics, at Carnegie Mellon, he has worked with Google, Qualcomm, and Honda Engineering, in application areas as diverse as augmented reality and self-learning binpicking systems. He now holds several industrial and academic patents, and has received several awards for his academic excellence. With a strong go-getter attitude, and solid knowledge of the latest trends in computer vision, artificial inteligence, convex optimization, and differential geometry, he is excited to tackle the most challenging robot-perception problems MUJIN has to face.
Yangjing Li is our resident electrical engineer. After getting a dual degree in Electrical Engineering and Art History, she received master's degree in Medical Device Design from Peking University. When Yurki is not designing circuits or tinkering with her latest gadgets, she enjoys martial arts and theaters.
System Integration Engineer
Robert is Mujin's system integration engineer making sure all delivered Mujin systems meet manufacturing specifications and are easy to use. Originally from Beijing, he was the first prize winner of the national Mathematical Olympiad out of tens of thousands competitors. He then studied electrical, mechanical, and industrial engineering from Hong Kong University of Science and Technology and University of Toronto. After Master's, he became the lead developer at a well-established robot system integrator. Now at Mujin, Robert wishes help bring the new technology revolution to reality.
Takeo Kanade is a professor of Computer Science and Robotics at Carnegie Mellon. Professor Kanade has worked in multiple areas of robotics: computer vision, multimedia, manipulators, autonomous ground and air mobile robots, and sensors. He has written more than 300 technical papers and reports in these areas, as well as more than 20 patents. He has been the principal investigator of more than a dozen major vision and robotics projects at Carnegie Mellon. Professor Kanade has been elected to the National Academy of Engineering. He is a fellow of the IEEE, the ACM, and American Association of Artificial Intelligence (AAAI), and the former and founding editor of International Journal of Computer Vision. Professor Kanade has served for government, industry, and university advisory or consultant committees, including Aeronautics and Space Engineering Board (ASEB) of National Research Council, NASA's Advanced Technology Advisory Committee, and PITAC Panel for Transforming Healthcare Panel.
Masayuki Inaba is a professor of Mechanical Engineering and Intelligent Systems at University of Tokyo. He is the head of the JSK Robotics Lab, which hosts some of the brightest robotics students in Japan. His research projects have included hand-eye coordination in rope handling, vision-based robotic server system, remote-brained robot approach, whole-body behaviors in humanoids, robot sensor suit with electrically conductive fabric, flexible-spine humanoid and developmental JSK mother projects with the remote-brained system environment, life-size assistive humanoids, musculo-skeletal spine humanoid series, whole-body soft sensor tissues, IRT home assistance with personal mobility, open-source robotics middlewares, high speed-and-powered legs for the next generation, on the budget from MEXT and MITI.
Phạm Quang Cường was born in Hanoi, Vietnam. He graduated from École Normale Supérieure rue d'Ulm (France) in 2007 and obtained a PhD in Neuroscience from Université Paris VI and Collège de France in 2009. From 2010 to 2013 he was a researcher at University of São Paulo (Brazil) and at University of Tokyo (Japan). He joined the School of Mechanical and Aerospace Engineering, NTU (Singapore) as an Assistant Professor in 2013.
We'd love to hear from you
Mujin Inc. was founded in 2011 in Tokyo, Japan by Issei Takino and Dr. Rosen Diankov, founder and active maintainer of OpenRAVE, the opensource platform used by hundreds of researchers to test, develop, and deploy motion planning algorithms for robotic applications. The team at Mujin is now working with robot makers, manufacturing engineers, and research institutions around the world to push industrial robotics and factory automation (FA) with fast, efficient, and general solutions that will revolutionize manufacturing.
The standard system integration process of manually designing and testing ad-hoc manufacturing solutions is time consuming, inefficient, and no longer practical given the current rapidly changing markets. Present-day manufacturing requires automated solutions that eliminate the downtime incurred by development and deployment cycles, leverage the full potential of advanced robotic hardware, and increase the overall efficiency and performance of production. The Mujin team consists of experts who know how to apply the state-of-the-art robotic technologies. By providing a large range of solutions, from stand-alone platforms to fully integrated services, Mujin strives to make "industrial robots for everyone" a reality, and increase the productivity around the world.
3-31-3 Yushima, Bunkyo-ku, Tokyo, Japan
Yushima Tohou building 1F