In　recently　years,　driver　fatigue　detecting　system　has　gained　increasing　attentions　in　the　area　of　public　security.　Researchers　have　succeeded　in　applying　the　EEG　signals　to　accurately　detect　individuals　fatigue　state　in　sustained　attention　tasks.　However,　these　studies　were　performed　under　laboratory-oriented　configurations　using　tethered,　ponderous　EEG　equipment,　which　are　not　feasible　to　develop　the　fatigue　detecting　system　in　the　real　environment.　This　study　focused　on　developing　a　portable　attention　level　monitoring　and　alarming　(ALMA)　system,　featuring　a　mobile　NeuroSky　MindSet　and　an　android　pad　based　real-time　EEG　processing　platform,　for　the　driver　fatigue　in　the　pervasive　environment.　A　brain　feature　rule　which　can　represent　the　brain　gradual　process　from　focus　state　to　the　fatigue　state　has　been　formulated.　We　evaluated　the　ability　of　attention　level　of　the　system　in　the　simulated　driving　cockpit　and　demonstrated　that　the　system　can　classify　the　subjects　attention　level　in　accordance　with　the　rule　in　the　real　time.　©　Springer　International　Publishing　2013.