I am developing a game in Unity3d.

In this game, 2 teams will fight without any human interaction. The fight results must be always the same, if the same teams are the same. This way, I must reproduce the same fight in many devices, getting the same results in each device.

The map where the fight happens is a 3d environment, so, my soldiers can move in any direction. And the soldiers cannot walk in the same space of another soldier, so I need an agent avoidance algorithm.

I am having some trouble to find a solution to the problem above. To achieve the 'always the same results' I am treating this fight as Turn Based Strategy game, with little fixed time intervals between the turns.

To calculate the pathfind of the soldiers, I am using the Unity3d navmesh framework. And then I move manually my agents according to the 'NavMeshAgent' 'steeringTarget' property.

But this method doesn't cover the obstacle avoidance algorithm, and my agents overlaps.

So, I am looking for a solution to my problems, It should be a manual obstacle avoidance algorithm, or a way to unity navmesh work in fixed time intervals, or a completely new approach to my path-finding problems.

I am developing a game in Unity.

In this game, 2 teams will fight without any human interaction. The fight results must be always the same, if the same teams are the same. This way, I must reproduce the same fight in many devices, getting the same results in each device.

The map where the fight happens is a 3d environment, so, my soldiers can move in any direction. And the soldiers cannot walk in the same space of another soldier, so I need an agent avoidance algorithm.

I am having some trouble to find a solution to the problem above. To achieve the 'always the same results' I am treating this fight as Turn Based Strategy game, with little fixed time intervals between the turns.

To calculate the paths of the soldiers, I am using the Unity navmesh framework. And then I manually move my agents according to the 'NavMeshAgent' 'steeringTarget' property.

But this method doesn't cover the obstacle avoidance algorithm, and my agents overlap each other.

So, I am looking for a solution to my problems, It should be a manual obstacle avoidance algorithm, or a way for Unity's navmeshbto work in fixed time intervals, or a completely new approach to my path-finding problems.