In the MultiModal contribution, it is defined a very good way to retrieve the TravelTime for walk and bike by considering several factors, like age, gender and road slope factor. Specifically, for the first two parameters, the whole data are completely specified following the Weidmann values, whilst the slope factors should be extrapolated from the network. At the moment I could not find a network example where the slope tag is specified and moreover an explanation about how to extrapolate the slope factors from Open Street Map. I would appreciate if you could provide some insights about how to easily extrapolate such slope factors as well as how to embed them within the network input file.
we've done some work on using slopes in South Africa. I think the paper will appear at the ABMTrans conference in few months, but I can send you a draft copy if you want. At the last developer meeting in September 2016 we actually implemented 3D coordinates (not sure it is part of the latest release already). The default elevation is Double.NEGATIVE_INFINITY.
The method CoordUtils.calcEuclideanDistance(...) honours the elevation, but there is an alternative CoordUtils.calcProjectedEuclideanDistance(...) that only considers the 2D distance.
The one downside is that is you use 3D coordinates, you have to use it throughout the model (facilities, activities, etc.)
To get 3D coordinates, you can look into the playground.southafrica.utilities.coord3D.Utils3D class (I cache the SRTM tiles locally so that it doesn't consume so much bandwidth). Also, there is a conversion class playground.southafrica.utilities.network.ConvertMatsim2dToMatsim3d that can be of value.
PS: I believe the revised network writer checks for the presence of elevation, and writes it to file (and reads it accordingly) if it exists. So if there is no elevation, you won't see it in the network file.
Thank you very much for your quick reply and this is exactly what I was looking for. Anyway, if it is ok for you sending me a draft copy of your work, that would be really great. You can see my email in the profile description.