Maximising reward from a team of surveillance drones : a simheuristic approach to the stochastic team orienteering problem