We are researching and developing end-to-end infrastructures supporting a broad spectrum of Big Data applications in our select sport and medical domains. A key property is geo-distributed scalability into public clouds and edge nodes with low cost profiles. Projects include the following:
EONS: Efficient Execution of Large Workloads on Elastic Heterogeneous Resources; a NFR-funded FRINATEK project investigating low-level systems aspects of Big Data infrastructures with focus on parallel programming and processing in the context of future distributed large-scale heterogeneous systems.
LoNet; a project researching and developing infrastructures that enable implementation of privacy-preserving enforcement and auditing. Privacy is a first order design concern, and our goal is to devise an overall architecture and develop a series of associated robust, privacy-preserving middleware systems supporting sharing and use of private data without leaking individual information.
DIGGI; a project researching and developing a generic IoT/mobile/cloud middleware platform for trustworthy computing. This is a platform where IoT devices, mobiles, smart home and personal computers, hybrid cloud solutions, and proprietary cloud solutions from different vendors seamlessly can be connected and integrated in a privacy-preserving and secure manner. Secure enclave technologies (Intel SGX and ARM TrustZone) provide the hardware foundations for this project.
We conjecture that new types of security mechanisms are needed when data no longer is stored in one or a few centralized stores. This is a rapidly emerging scenario, where individuals already today have data scattered around in the cloud, in locked-in web services, in social networks, and on cellular phones, USB-sticks, digital cameras, wearables, personal computers, and enterprise servers. Add to this the emerging new sensor streams and embedded devices capturing personal user data 24/7 in the proximity of a modern citizen.