This　study　presents　the　reliability　and　inspection　optimization　of　a　novel　k-out-of-n:　G　system　with　linear　consecutive　k　＇-out-of-n　＇　load-sharing　subsystems,　abbreviated　as　k-out-of-n/Cl(k　＇,n　＇;F),　motivated　by　a　roller　system.　The　subsystems　are　identical　and　mutually　independent　and　the　system　works　when　at　least　k　subsystems　function.　Within　a　subsystem,　failure　dependence　among　components　exists　since　the　carried　load　is　unequally　shared　by　surviving　components.　We　model　the　load-sharing　effect　by　the　tempered　failure　rate　model.　The　failure　sequence　diagram　(FSD)　is　constructed　to　describe　possible　events　of　a　subsystem　evolution　in　the　present　of　the　states　of　components.　A　regular　inspection　policy　is　proposed,　upon　which　hidden　failures　at　the　component　level　can　be　revealed.　The　system　is　preventively　maintained　upon　an　inspection　or　at　its　failure,　which　occurs　first.　The　long-run　expected　cost　per　unit　time　is　minimized,　and　a　coupling　search　FSD　and　sampling　algorithm　is　designed　to　search　for　the　optimal　inspection　interval.　A　case　study　is　presented　to　show　the　applicability　of　the　proposed　model,　where　we　show　the　reliability　variation　and　indicate　that　the　current　one-week-inspection　policy　is　too　conservative.　Sensitivity　analysis　is　provided　to　examine　the　impact　of　cost　parameters　on　the　optimal　decision.