Robot fleet manager from open-source software

Desired outcome

When fleets of robots are deployed, in say, a warehouse they require additional systems to deconflict the routing of different robots, as well as dispatch robots on specific missions and provide other services. Traditionally, this has been done with expensive commercial offerings. However, there now exist open sourced solutions that provide much of the needed capabilities. However - as is common with open source software - this needs to be packaged with other components and set up for a particular scenario to be useful. This project would bring all these elements together to create more useful and usable software distribution that can be used to control fleets of robots. Think of this project as being the Ubuntu of Robot software.

Initial Problem Description

The ultimate goal is to enable robot users all over the world to use large fleets of robots easily.
Currently they must used closed source offerings which are expensive, and over-which they have no control. This is terrible as they usually need to integrate their robot fleet with a warehouse management system and so lose control of their business to the software suppliers.
The desired state is that we have a fleet manager, where large portions of the code are open and they can review and make changes. The specific parts that need to be open revolve around mission dispatch.
The problem statement is - how can we package up a set of open-source software with other Ubiquity Robotics offerings to make a software distribution that can easily be used to control large fleets of robots.

Context

With the rise in E-commerce, fulfillment warehouses have grown faster than the rest of the economy. Meanwhile labor cost increases and a decreasing interest in working manual labor jobs has meant that its become more difficult to recruit people to work in those warehouse jobs. 80% of hiring managers have had difficulty in recruiting post-pandemic. Most of the work in warehouses is material movement - often being 85% or more of the activity. Mobile robots are well positioned to take on that activity and there has been dramatic growth of robot usage in the largest warehouses. Ubiquity Robotics has developed a warehouse conveyance product that fits nicely in to this space. However, our current system is designed to operate by using advanced collision avoidance that sidesteps the need for a fleet manager.

Some customers like it, but our other, larger customers want a system with all the sophistication that a fleet manager brings. We know open source software can get us 95% of the way there - what is needed is to package the open source software together so that our customer needs can be met.

Connection to cross-cutting areas

These robots are used in warehouses, and in factories to move goods around. They are a prime example of Industry 4.0 and digitalisation in action. Their deployment help warehouses and factories become significantly more efficient while removing drudgery from the lives of warehouse workers - while at the same time closing the gap between the number of people needed for these jobs and the number of people available. They are an electric vehicle that displaces gas powered forklifts in warehouses so there is some connection to sustainability topics.

Input

There is extensive open-source software that can be leveraged. Specifically
Open RMF is an open source fleet manager that provides mission dispatch, route deconfliction and fleet planning https://www.open-rmf.org/ we are a consortium member of this piece of software.
Vizanti http://wiki.ros.org/vizanti is an open sourced robot control software - it was developed by an Ex-Ubiquity Robotics team member and provide visualizations, and planning. We also have an internal piece of software that does similar things called EZ-map.
ROSE - is a Ubiquity Robotics secure connectivity tool that allows users to connect to the robot securely and perform command line tasks on the robot.
LaMa SLAM -https://github.com/iris-ua/iris_lama_ros is an open sourced and high efficiency SLAM stack that may be useful.
There are many other both internal Ubiquity and open source tools and pieces of software that may be useful that can be discussed during the project kick off.

Expectations

We would like the team to deliver a distribution that can run on a simple server (we often use Raspberry Pi devices) that will enable a number of fleet management systems. The ideal situation would be that we enable a small robot fleet in our lab in Ljubljana and that you are able to experiment with this robot fleet. For this to be a reality you would need to start with

a) Remote software update capabilities (already exists with ROSE but you would need to standardize)
because you need a) to be able to remotely experiment with the robot fleet you would then move on to.

b) Remote robot control / driving capabilities (Vizanti / EZ-map)
c) Fleet planning (Open RMF)
d) Mission dispatch (Open RMF)
e) Robot health monitoring and analysis (Open RMF plus dashboards)

Desired Team Profile

This is a largely technical challenge - what is needed is a highly technical team to tackle it. The team would need a mixture of Java ./ front end focused students along side those who are heavily in to security, server side. Robot enthusiasts would always be welcome. People who want to understand robots are the people who would thrive here.

Additional Information

The competitors are companies like inOrbit - they make great offerings. However they are also expensive to install and use. Most importantly customers need to lock themselves in with inOrbit and tie their business processes to them. This is something that customers hate and also something that is counterproductive to the community as a whole. Using open-source tools would be really interesting and is possible in a short space of time.

Related Keywords

• Automation, Telepresence & Robotics
• Robotics