This　article　introduces　a　secure　tunnel　fast　marching　tree　motion　planning　algorithm　(ST-FMT*)　to　provide　a　secure　and　optimal　path　quickly　for　a　mobile　robot.　The　proposed　ST-FMT*　consists　of　preprocessing　and　exploring　procedures,　which　are　responsible　for　establishing　a　secure　tunnel　and　optimizing　the　path,　respectively.　In　the　preprocessing　process,　the　generalized　Voronoi　graph　is　adopted　to　build　an　equidistant　roadmap　and　generates　an　initial　collision-free　solution　rapidly.　Then,　a　secure　tunnel　is　established　via　the　minimum　distance　from　the　obstacles　to　the　initial　solution　to　facilitate　the　concentration　of　sampling.　In　the　exploration　process,　the　FMT*　with　uniform　sampling　within　the　secure　tunnel　is　utilized　to　find　the　optimal　solution　with　high　computational　efficiency.　The　theoretical　analyses　of　the　ST-FMT*　are　provided.　In　a　series　of　scenarios　evaluation,　the　ST-FMT*　exhibits　fast　convergence　to　the　optimal　solution　in　different　environments.　Besides,　the　effectiveness　of　the　ST-FMT*　is　verified　in　a　transportation　experiment　using　a　Turtlebot2　mobile　robot.