Fig. 2.

Overview of the basic growth algorithm in ConFiG. In this example, three fibres are shown with a growth network that only contains relevant nodes for the sake of visualisation. From the set of nodes, a network is constructed using the Delaunay triangulation. Each fibre then grows from node to node, along any edge connected to the current node. The node moved to will be the node with the lowest cost. Once a fibre segment has grown, the network nodes are updated to store information about which nodes are occupied or near to an existing fibres. This contributes to the cost function for any future fibres, penalising moving to nodes too close to existing fibres. It is not possible to move to any node now inside a fibre as indicated by the removal of this edges from the network (pairs of blue arrows show where this is happening). The next fibres grow, now avoiding existing fibres until all fibres have finished. See Supplementary Video 1 for an animation of this algorithm.