1000 robots or 1 (for one single unattended robot license)

I would need some help to understand how the license of unattended robot works.

STATUS QUO
Let’s say I have 1000 clients all over the world.
I would need from them to send me their inventory stock of 3 products (ex: 1250 x Item1, 7200 x Item2, 3100 x Item3), let’s say 10 times (tasks) per day. Each task will take no more then 5 seconds (they will copy/paste info from a desktop app in a web page).

QUESTION
Can we deploy one unattended robot on each client’s computer, so 1.000 unattended robots… which are sharing 1 license / runtime (unattended robot)? I mean all robots will wait in line to get the ONE runtime/license.

Is it this achievable with one single unattented robot license (may be 5 or 10, but 1 emphasize the ideea)?
I’m looking to figure out how this license could be used in such a project. May you point me out there to a project/video with such implementation?!

Obviously, we are looking to avoid one single unattented robot working with 1000 x Remote Desktop Connections which will eat significant more time and resources. We’re looking also to avoid colision with local workers on the local computers (running in background with RDC is not the solution since the desktop installed programms are so different from one client to another). That’s why we’re looking to deploy local unattented robots.

Cheers

Basically what you are looking for is Floating robot…

Following thread to understand if possible how that can be achieved for knowledge :slight_smile:

@loginerror can you please take a look at the question above

That page explain nothing, Shubham.

Regarding the example UiPath provided here, it seems that my scenario is plausible:

For example, if on a machine with 5 Robots, you assign only 3 runtimes, it means only 3 licenses are consumed. If you execute 5 jobs or schedule all the Robots on that machine, the following occurs:

  1. 3 jobs are executed on the first 3 licensed Robots;
  2. the remaining 2 unlicensed Robots are placed in a pending state;
  3. the first 3 Robots release their licenses;
  4. the remaining 2 Robots are licensed and they execute the assigned jobs.

I would like to have a confirmation from someone which implemented already such scenario. Or is creating a project based on this assumption.

How are your 1000 clients going to connect to your UiPath Orchestrator system?

The number of Robots isn’t really relevant. Licenses and Robots aren’t 1-to-1.

Is Orchestrator limited in some way to not accommodate 1000 unattended robots / clients / employees or remote computers?

No. Are you going to have a VPN set up? Cloud version? If you only have one license only one of your 1000 clients will be able to connect at a time. This will be very difficult to manage. 50 seconds per day, for 1000 clients, is 50000 seconds which is almost 14 hours.

If they’re just sending you info on stock numbers, there is probably a better way for them to do that.

I have to second that. Managing the robot installation itself on so many different environments would be a hassle.

One idea to handle such use case could be UiPath Apps. You could create an app that allows anyone with access to input and submit their data. That app would then be connected to an unattended robot on your backend to process it. If multiple customers user the app at the same time, I think it would simply queue the jobs without an issue.

The only thing to manage here would be access to the app itself.

You wouldn’t even have to use UiPath Apps. You could just publish a web page that emails formatted data to a mailbox the bot picks it up from. Or even submits directly to the Orchestrator Queue using the Orchestrator API.

1 Like

Thanks for your ideas.

Main issue is that I would need to NOT involve any human action on the client side. That’s why I am looking for a solution to automate locally the extract and send data.

Moreover, there are some 100 different systems where these data stock sit and I don’t have access to their database, only through UI.

With UIPath and Computer Vision I can “teach” robots to extract those data much easier then developing an app to accommodate all these different local softwares etc.