build details

Show: section status errors & todos local changes recent changes last change in-page changes feedback controls

Challenge AMOD

Modified 2019-04-17 by clruch

We introduce a challenge that is presently too extensive to conduct in physical Duckietown. Instead, the task is carried out in simulation. In the future, this is expected change as Duckietowns across the world experience a tremendous growth in population, mostly due to the significant immigration of Duckies with an engineering background.

We use the simulation platform AMoDeus AMoDEUs [6]. AMoDeus is a virtual mobility-on-demand simulator on which operational policies for large-scale AMoD systems can be developed and tested on a high fidelity mesoscopic transportation simulator. You may coordinate thousands of robotic taxis in cities to pick up and deliver customers with your algorithm hosted by the AMoDeus engine in the background.

The design of operational policies for AMoD systems is a challenging task for various reasons:

  • The vehicles operate on a large and complex network with varying travel-times and time-varying congestion effects.

  • Every decision which is taken influences most of the other decisions. For instance, if a taxi has picked up a customer, the vehicle is not available to service other close-by requests until the customer is dropped off.

  • Running an AMoD system with good service levels is a task that requires to chose the number of employed robotic taxis correctly, to reduce the empty miles driven and maintain good wait and journey times for the customers. These are conflicting goals.

Therefore, operating AMoD systems is a task that requires complex decision-making and anticipative action-taking. This makes it an ideal duckietown-challenge!

Development guide

Modified 2019-04-14 by Liam Paull

No questions found. You can ask a question on the website.