Tower　crane　scheduling　is　a　classic　conundrum.　It　is　further　complicated　to　prevent　collisions　and　reduce　idle　transportation　time　among　multiple　overlapping　tower　cranes.　Unlike　previous　research　attempts,　this　study　aims　to　provide　an　optimal　solution　to　this　multiple　crane　service　scheduling　problem　(MCSSP).　Firstly,　the　MCSSP　was　translated　to　a　Mixed　Integer　Linear　Programming　(MILP)　model.　Then,　the　model　was　optimized　by　1)　distributing　the　lifting　requests　in　overlapping　areas　to　the　proper　tower　cranes;　2)　selecting　the　appropriate　supply　location　to　serve　each　lifting　request;　and　3)　arranging　the　lifting　sequences　of　each　tower　crane　to　complete　the　requests.　Compared　with　previous　methods,　the　proposed　MILP　model　(solved　using　Gurobi™)　can　result　in　the　saving　of　6.54%–18.07%　of　total　operation　costs　meanwhile　achieving　the　non-collision　goal.　The　findings　of　this　research　can　be　deployed　in　optimizing　efficiency　and　safety　in　the　real-life　scheduling　of　multiple　overlapping　tower　cranes.　©　2021　Elsevier　B.V.