Building
Robot Cloud Portals
Made Easy
Transitive solves the make vs. buy dilemma of robot fleet management systems by giving you the full-stack robotic capabilities you need to build your own.
Built on a novel, open-source framework for full-stack robotics.
Or chat with us on Slack.
What are Capabilities?
Capabilities are independent modules each consisting of a robot component, a cloud component, and one or more front-end/web components. These components communicate seamlessly through the Transitive framework, working together to add new, interactive functionalities to your fleet.
The Transitive framework is specifically designed to allow these web-components to be embedded in your own web portals and pages, just like embedding a YouTube video but with authentication to provide access control. Each capability is individually priced and you only pay for the capabilities you are actively using.
Several of our capabilities support ROS (1 + 2) and work well with it, but Transitive itself does not require ROS.
Example Capability: Remote Teleop
Install the agent
Install the Transitive agent on your robot — sudo not required! — or in your docker image. The agent is responsible for installing capabilities and running them in designated sandbox environments.
See your robots online, install capabilities
The agent connects to the Transitive cloud via Transitive's data-sync protocol which extends MQTT for safely synchronizing data. Once connected, you can see your robot on the Transitive Portal. From here you can install capabilities, just like adding apps to your phone. Alternatively, when using docker, you can have capabilities pre-installed in the image.
Test and configure capabilities
Once a capability is installed and running, you can quickly test it out on the Transitive Portal via its UI components. Some capabilities, such as remote-teleop, also let you pre-configure UI components from here.
Embed the configured UI element in your own web portal
When you are satisfied with your configuration, open the Embed modal which shows you how to embed the pre-configured component in your own web portal or share it as a stand-alone page. Authentication and authorization is solved via a JSON Web Token (JWT) that you sign with your Transitive JWT secret, available in your account.
Enjoy the new capability in your portal
The capability you configured is now available in your web portal and ready for use. All required connections are handled by Transitive, incl. cloud-proxies and TURN server as required by the capabilities you install. No development risk, no maintenance overhead, and all future updates and necessary security patches happen automatically and are included in the monthly fee for the capability.
Make vs. Buy: solved
Every robotics company faces this dilemma: should we build our own fleet management system or use a third-party SaaS offering? Neither of these options is good because the former takes a long-time, defocuses the company, and takes a lot of money. The latter suffers from a glass ceiling: eventually the capabilities of those third party solutions will not meet our needs anymore because every robotics application and company is different. This is why we built Transitive, to create a middle-ground between "make" and "buy": making it easy to build your own, perfect fleet management system by giving you the modules you needs.
Christian created the open-source Transitive framework out of the belief that the robotics industry is lacking a shared, open platform for the development of full-stack robotic software. As a result, robotics companies lose time creating their own cloud middle-ware for robotics and reimplementing dozens of common full-stack components when this time could be better spent on their specific product. Christian has worked for six different robotics startups and has seen this time and time again. Previously, Christian was VP of Software at Savioke, maker of the Relay delivery robot for hotels and hospitals, where he helped scale the fleet from 5 to over 100 robots operating on three continents. Christian earned a PhD in artificial intelligence from the University of Toronto and is the author of over 40 publications and 17 patents.