In　the　mobile　industrial　human　machine　interaction　(HMI),　the　engineer　has　to　manually　select　the　target　machine　from　a　list　to　establish　the　data　connection.　It　is　a　nontrivial　problem　since　there　are　more　machines　connected　to　the　industrial　cyber-physical　systems,　and　the　list　becomes　so　long　that　is　hard　to　be　read.　Considering　the　characteristic　that　the　industrial　HMI　is　typically　executed　in　a　face-to-machine　manner,　the　face-to-machine　proximity　estimation　(FaceME)　algorithm　has　been　proposed　to　simplify　the　manual　selection.　However,　due　to　the　limited　utilization　of　RSSI　data,　the　estimation　accuracy　of　FaceME　is　not　sufficient　in　the　application　with　densely　deployed　machines.　In　this　paper,　we　propose　the　model　to　formulate　the　neighbor　relations,　and　then　design　a　new　face-to-machine　proximity　estimation　algorithm　with　neighbor　relations　(FaceMEN)　to　improve　the　estimation　accuracy.　The　experimental　results　prove　that　FaceMEN　greatly　improves　the　estimation　accuracy　and　time　complexity.　©　2018　IEEE.