A　collaborative　truck-and-drone　system　(TDS)　can　perform　various　tasks,　such　as　military　surveillance,　recon-naissance,　logistic　delivery,　disaster　search　or　rescue.　In　order　to　enhance　the　survivability　of　such　a　system　and　improve　the　probability　of　task　success,　the　task　can　be　aborted　and　a　rescue　procedure　can　then　be　activated　when　a　certain　condition　relating　to　malfunction　or　incident　management　is　satisfied.　Multiple　drones　can　work　together　to　complete　a　task　with　high　reliability　once　a　single　drone　is　unable　to　respond　to　complicated　emergencies.　To　consider　this　challenge,　this　paper　designs　a　joint　optimisation　model　to　consider　task　abortion　when　routes　of　trucks　and　drone　cluster　are　assumed　under　random　attacks.　Additionally,　the　paper　considers　time　windows　of　targets　and　the　range　of　the　truck　for　protecting　drones　in　the　routines　of　a　TDS.　To　minimise　the　expected　total　cost　due　to　trucks＇　destruction,　drones＇　destruction　and　unvisited　targets,　we　obtain　the　optimal　truck-and-drone　routing　strategy.　Some　numerical　examples　on　Solomon　datasets　are　given　to　illustrate　the　applicability　of　the　proposed　abortion　strategy,　present　the　results　of　sensitivity　analysis　on　the　drone　cluster,　and　then　prove　the　effectiveness　of　the　optimisation　method.