This　study　proposes　an　interesting　customized　bus　service　design　problem　by　considering　travel　demand　uncertainty.　Given　a　fleet　of　heterogeneous　vehicles,　a　mixed　integer　linear　programming　(MILP)　model　is　put　forward　for　the　complex　decision　making　on　bus　routing,　timetabling　and　bus　deployment,　with　the　objective　of　generating　a　set　of　profitable　bus　services　to　cater　for　diverse　commuting-trip　requests.　To　capture　the　risk-averse　level　of　the　bus　operator　in　uncertain　travel　demand　environment,　a　random　variable　describing　the　likelihood　that　the　offered　bus　services　are　rejected　by　potential　passengers　and　two　associated　control　parameters　are　embedded　in　the　MILP　model,　facilitating　an　adjustable　robust　optimization　framework.　A　branch-and-price　method　is　implemented　to　solve　the　model　exactly.　A　column-generation-based　heuristic　method　is　proposed　to　solve　large-scale　problems.　The　effectiveness　of　both　the　exact　and　heuristic　methods　is　assessed　in　numerical　experiments.　©　2020　Hong　Kong　Society　for　Transportation　Studies　Limited.