Variable　polarity　plasma　arc　welding　(VPPAW)　is　a　highly　efficient　method　for　joining　aluminum　and　magnesium　alloys,　primarily　because　of　the　automatic　removal　of　oxide　layers　through　cathode　spots.　Additionally,　the　high　energy　density　generated　by　a　variable　polarity　plasma　arc　creates　a　keyhole　within　the　weld　pool,　enabling　single-pass　welding　of　medium-thick　plates　with　minimal　deformation　and　residual　stress.　Understanding　the　keyhole　behaviors,　arc　physics,　and　control　methodologies　are　essential　to　optimize　and　improve　the　stability　of　this　welding　process.　This　survey　paper　offers　an　overview　of　the　development　of　the　VPPAW　process　and　its　recent　developments　of　detecting,　modeling,　and　controlling.　It　shows　that　the　potential　applications　of　hybrid　processes　derived　from　VPPAW　have　expanded.　Advanced　process　detection　techniques,　particularly　the　utilization　of　X-ray-based　vision　systems　for　the　in-situ　observation　of　molten　metal　flow,　have　significantly　contributed　to　the　understanding　of　VPPAW.　Moreover,　studies　investigating　VPPAW　across　different　welding　positions　have　enhanced　the　adaptability　of　this　welding　method.　Accurate　numerical　models　and　intelligent　deep　learning　methodologies　have　been　crucial　in　unveiling　the　underlying　mechanisms　and　facilitating　the　predictive　analysis　of　welding　quality.