Statistical reconciliation of seasonally adjusted time series

Dear all, I have launched this project in order to find a solution to the following statistical/ mathematical problem. In time series analysis there exists a problem of reconciliation of time series. It is especially pronounced for seasonally adjusted time series. When time series are directly seasonally adjusted the component series do not add up to the total anymore. This is problematic for data users and statistical offices as they prefer data consistency. Therefore, the solution to the mathematical problem has to be found with the set of constraints, temporal and contemporaneous. The quality of reconciliation has to be assessed so that it proves to be a reasonable procedure to be carried out. In the end, a routine in some software has to be offered for reconciling times series. In the end assessment of reconciliation as well as of seasonal adjustment quality has to be conducted. Seasonal adjustment quality needs not to be harmed in an excessive way. There has to be no major change to the series level. There exists a number of methods already (Fortier (2012), Di Fonzo and Marini (2005 and 2011), prorating or raking). A more simple and handy software implementation is appreciated. Therefore, I am looking forward to your comments and suggested methods and will be glad to discuss any issues around the topic and possibilities for the cooperation