Industry　4.0　intends　to　realize　mass　individualization　production　driven　by　customers＇　demands,　thus　generating　several　advanced　manufacturing　modes.　However,　the　existing　modes　have　disadvantages　such　as　low　customer　participation,　low　response　for　modifications　in　real-time,　and　low　utilization　of　distributed　idle　productivity　resources.　The　above　deficiencies　would　not　cater　to　the　idea　of　Industry　4.0　well,　making　the　cost　of　individualization　products　too　high　or　the　degree　of　individualization　too　low.　In　this　paper,　a　customer　highly　involved　dynamic　demand-driven　smart　manufacturing　mode　and　a　mathematical　model　are　presented　for　mass　individualization　production.　Under　this　mode,　customers　can　be　prosumers,　participating　in　the　whole　production　process　and　proposing　modifications.　Meanwhile,　multiple　distributed　idle　productivity　resources　are　employed　with　dynamic　demand-driven　manufacturing.　In　consequence　of　the　mathematical　model　and　algorithms,　the　multiple　task　allocation　among　multiple　factories　problem　is　solved,　maximizing　the　total　profit　of　available　factories　with　high　efficiency　and　low　resource　waste.　Experiments　and　comparisons　are　made　about　the　feasibility　and　efficiency　of　the　algorithms.　©　2021　IEEE.