- Why use Innovation Place
- Funds for r&d
- About us
- PNO services
Synthetic data is artificially manufactured data that mimics the original data.
It keeps the statistical distribution as the original data but doesn’t contain sensitive information and can thus be shared internally and externally for further analysis.
Synthetic data is part of Privacy-Enhancing Technologies (PET). Like previous waves of cryptography, PET’s adoption could unlock a trillion dollars opportunity by helping extract more value from existing data, driving the creation of even more of it and enabling a new generation of services and use-cases.
It removes Personally Identifiable Information (PII) and thus is considered being fully anonymized. Synthetic data cannot be reverse engineered as it is the case with pseudonymisation. It allows to be fully compliant with existing regulations and to prevent from potential fines and penalties. It can be scaled to any size and sampled to an unlimited extent, making it highly efficient.
Synthetic data can be applied by anyone handling sensitive data and the fields of application are manifold, such as
- company-wide sharing and artificial intelligence (AI) trainings for company internal sharing
- cross-company sharing and collaborative research for external sharing
Overview of fields of application and synthetic data generator, see pictures below.
The Swiss company's background is in data science and it offers services that use deep learning to generate synthetic data for various file formats. The clients are institutions facing challenges such as compliance laws, fear of data misuse, patient/customer privacy, etc.
The company has identified the following topics in Horizon Europe in which they could provide its expertise and become a valuable partner for a consortium:
TOPIC ID: HORIZON-CL3-2021-CS-01-04
TOPIC ID: HORIZON-CL4-2021-DATA-01-01
TOPIC ID: HORIZON-CL4-2021-DATA-01-03
TOPIC ID: EDF-2021-PROTMOB-D-DMM
TOPIC ID: EDF-2021-OPEN-R-SME
TOPIC ID: EDF-2021-OPEN-RDIS-Open
TOPIC ID: HORIZON-HLTH-2022-TOOL-11-02
TOPIC ID: HORIZON-HLTH-2022-IND-13-02
TOPIC ID: HORIZON-CL6-2022-GOVERNANCE-01-10
TOPIC ID: HORIZON-CL6-2022-GOVERNANCE-01-11
TOPIC ID: HORIZON-INFRA-2022-EOSC-01-03
It is also looking for industrial partners in the health, aerospace, security and defense or manufacturing industry to identify use cases under commercial agreements with technical assistance.
Advantages & innovations
Synthetic data is part of Privacy-Enhancing Technologies (PET). In comparison with other PET’s, synthetic data doesn’t encrypt the data and therefore there’s no issues around a key. Frequently existing business problems such as data privacy or not enough data for effective AI (artificial intelligence) training can be solved by using synthetic data in an innovative way. The Swiss company has developed and implemented a deep learning model for creating synthetic tabular data on a free basis. The solution cannot be reversed and is: • fully compliant with privacy regulations • easy and flexible to synthesize tabular data on a free basis • cost effective when compared to alternative solutions • capable of scaling to commercial level needs in a very short amount of time
Stage of development
Already on the market
Research Cooperation Agreement: - Coordinators of a Horizon Europe proposal such as: TOPIC ID: HORIZON-CL3-2021-CS-01-04 TOPIC ID: HORIZON-CL4-2021-DATA-01-01 TOPIC ID: HORIZON-CL4-2021-DATA-01-03 TOPIC ID: EDF-2021-PROTMOB-D-DMM TOPIC ID: EDF-2021-OPEN-R-SME TOPIC ID: EDF-2021-OPEN-RDIS-Open TOPIC ID: HORIZON-HLTH-2022-TOOL-11-02 TOPIC ID: HORIZON-HLTH-2022-IND-13-02 TOPIC ID: HORIZON-CL6-2022-GOVERNANCE-01-10 TOPIC ID: HORIZON-CL6-2022-GOVERNANCE-01-11 TOPIC ID: HORIZON-INFRA-2022-EOSC-01-03 Commercial agreements with technical assistance: - industry partners from health, aerospace, security and defense or manufacturing The tasks to be performed by the partner sought: State and define an internal challenge concerning data sharing such as: - cannot share data internally/externally because of privacy regulations - cannot perform data analytics because of lack of data - partial/no access to data A use case for implementing synthetic data is then derived from the definition of the challenges.
Source/ contact: Enterprise Europe Network (europa.eu)
IF YOU ARE INTERESTED IN KNOWING MORE ABOUT THIS PROJECT PLEASE REGISTER AS FREE MEMBER OR LOGIN IF ALREADY REGISTERED