The　cloud　robotic　technology　makes　multiple　robots　share　resources　and　collaborate　with　each　other　more　flexibly.　Although　this　concept　has　been　widely　accepted　by　academia　since　it　was　put　forward,　it　still　faces　many　challenges　in　engineering.　The　main　problems　are　that　the　functions　of　industrial　robotic　systems　become　more　and　more　complex,　and　the　programming　language　and　computing　environment　of　multiple　vendor　robots　are　significantly　different,　accordingly　the　cooperative　control　of　different　devices　and　the　online　testing　become　extremely　difficult.　Therefore,　this　paper　proposes　a　new　control　platform　for　cloud　robots,　called　IAPcloud,　which　has　significant　advantages　in　solving　the　problem　of　collaborative　control　among　heterogeneous　robots　and　their　auxiliary　devices,　declining　the　programming　difficulties　of　cloud　robotic　control　systems　and　shortening　the　development　and　deployment　cycle　of　the　application.　Finally,　we　verify　the　scientific　and　effectiveness　of　the　IAPcloud　platform　by　three　cases,　including　the　image　recognition　and　tracking,　the　human-machine　gobang　game,　and　the　online　dynamic　reconfiguration　of　control　algorithms.　©　2013　IEEE.