Multidisciplinary　Design　Optimization　(MDO)　is　proposed　for　the　design　of　the　lines　of　an　underwater　robot.　Hydrodynamic　performances　of　the　underwater　robot　are　concerned　about　in　the　design,　including　the　resistance　and　the　manoeuvrability.　A　method　of　MDO,　Collaborative　Optimization　(CO),　is　adopted.　To　improve　the　efficiency　of　optimization,　approximate　models　are　established　in　sub-disciplines.　Also,　an　artificial　intelligent　technique,　Particle　Swarm　Optimization　(PSO),　is　incorporated　into　the　CO　framework.　The　optimization　design　of　the　lines　of　the　underwater　robot　is　carried　out　on　an　Isight　platform,　an　automatic　integration　optimization　platform.　Through　the　platform,　the　optimization　design　of　the　lines　and　the　analysis　of　the　hydrodynamic　performances　can　be　achieved　automatically　with　high　efficiency.　An　autonomous　underwater　vehicle　(AUV),　SUBOFF,　is　taken　as　a　verification　model.　For　different　lines,　CFD　calculation　is　performed　to　analyze　the　resistance　and　manoeuvrability.　By　comparison,　the　optimal　lines　of　the　hull　and　the　fairwater　are　determined.　©　2015　Elsevier　Ltd.　All　rights　reserved.