Long-term sustainability mechanisms are needed to deal with technological obsolescence (e.g. of formats, software), “semantic decay” of digital assets (understanding changes over time, e.g. due to changes in use of terminology), and discipline and societal changes as the data evolves across different phases of the information lifecycle (socialisation of the data). Maintaining the complex relationships and dependencies between the components of this ecosystem is key to achieving “preservation by design”, through models that capture intents and interpretation contexts associated with digital content. This enables data to remain relevant to the intended and new communities of users. The volume of data and the cost of manual intervention means that human appraisal is highly labour-intensive and increasingly expensive, especially as content may be appraised on multiple occasions during the data lifecycle. PERICLES will address these challenges through the development of methods and tools that enable preservation systems to ensure volatile, heterogeneous data remains relevant to existing and new users through the management of the semantic, context and process (and policy) lifecycle. The project will also offer methods and tools to help formalise the appraisal process enabling automation and reproducibility of appraisal and reappraisal of data in a scalable manner. A number of domain-based communities of practice (in media, science and engineering) will be established and used to drive the development which will employ current state-of-the-art technologies (such as iRODS, intelligent Clouds).