The genesis of this work began during the DESURBS project. The scope of the project was to help build a collaborative decision-support system portal where spatial planning professionals could learn about designing much more secure and safer spaces in urban areas. The portal achieved this via integrating a number of tools under a common, simple to use, interface. However, the deficiencies in the project became apparent with subsequent development. Many of the open data employed changed format while applications were increasingly custom built for a single dataset. In order to overcome this a system called KnowDS was redesigned. The essence of the new design includes decoupling acquisition, analysis and overall presentation of data components. The acquisition component was designed to snap-shot the “data providing methods” and query data provenance in a similar way to a source code repository. The analysis component is built under a number of modular tools with a common interface which allows analysis to build in a plug&play approach. Finally, the data presentation component is where the custom logic goes. Under such design approach, the building of future applications becomes less challenging. As a consequence, two case studies using the new framework were considered. Firstly, a UK crime web-browser which allows data analytics performances at various granularities of crime types while correlating crimes across various UK cities has been achieved. Secondly, a mobile application which enables to generate reports on citizens’ perception of safety in urban zones has also been developed. The two applications were efficiently built under the new design framework; and they clearly demonstrate the capacity of the new system while they actively generate new knowledge about safety in urban spaces.