Revision history [back]

Hi @LuisCa

There is an interesting publication that used Heron robot and trained it with reinforcement learning. New worlds were created and you can use those to test Heron as well. Here is the link to the paper

The fork of this project can be found: https://github.com/AntoineRichard/Heron-RL-ICRA. However some of the links are not working, so you need to dig a little.

For example the uuv_simulator/uuv_gazebo_worlds/worlds/ has other worlds available for simulation for example lake.world.

Not sure about size, but you can modify them for sure.

Also take a look at the heron_worlds package that was made to help create world specifically for Heron: https://github.com/heron/heron_worlds

Hi @LuisCa

There is an interesting publication that used Heron robot and trained it with reinforcement learning. New worlds were created and you can use those to test Heron as well. Here is the link to the paper

The fork of this project can be found: https://github.com/AntoineRichard/Heron-RL-ICRA. However some of the links are not working, so you need to dig a little.

For example the uuv_simulator/uuv_gazebo_worlds/worlds/ has other worlds available for simulation for example lake.world.: https://github.com/AntoineRichard/uuv_simulator/tree/master/uuv_gazebo_worlds/worlds

Not sure about size, but you can modify them for sure.

Also take a look at the heron_worlds package that was made to help create world specifically for Heron: https://github.com/heron/heron_worlds