I must say "not good".
Main problem is time.
Another problem is that I'm probably overestimating my abilities. So this creates some fear of me failing again and thus lovers my interest to continue.

But all is not lost!
I might just go and try to make a different idea. The final decision on this will be made after I manage to implement a modified A* pathfinding algorithm.
Since for both ideas I require a modified A* pathfinding algorithm where I could take into account athleast two parameters defining "weight" vale for traveling from one node to the other (relative distance, path safety).

So since I'm reimplementing A* algorithm I was thinking of also adding heuristic support to it. While this is actually not needed for my first idea it would come verry handy for my second one.


My second idea is to make a City traffic simulator where "protagonists" would be cars driving around the city and player would have the ability to manage traffic by:
- building roads (extending roads with existing lanes)
- extending crossings with "turn lanes"
- equiping the crossings with "traffic lights" and setting the way how these "traffic lights" work (defining sequence)
- defining traffic direction by setting traffic direction for each lane
- defining speed limits for each lane
- reserving lanes for only certaing groups of vehicles
- maybe even adding pedestrian lanes (would double the needed time and work)
Cars would chose their route based on several parameters (hence multiparameter implementation of A*) like road section lenght, average speed for that section during previous day, type of road, etc.
Also for this idea I would use heuristic pathfinding where:
- on first heuristic level I would search posible paths between different road crossings
- on second heuristic level I would be doing pathfinding inside each road section (road between two road crossings) so cars would be able to change lanes and thus overtake stoped cars for instance (cars can break).
I would naturally have to implement some basic traffic rules defining which car can go through crossing first.
I would like to limit the field of view for cars to simulate one driver missing up another car. Combining this with simulating athleast basic vehicle physics (acceleration and deceleration) so it would be posible for two cars to actually crash one into another.


I belive this idea might offer more enjoyment to potential players as my first idea would be only suitable for "hardcore" simulation lovers.