CSE 5001 Research Seminar in Software and Service

  • Slides: 7
Download presentation
CS-E 5001 Research Seminar in Software and Service Engineering Fall 2016: Blockchains Project Topic:

CS-E 5001 Research Seminar in Software and Service Engineering Fall 2016: Blockchains Project Topic: Multi-Tenant Data Governance via trusted immutable ledger of shared data

Problem • Many industrial systems consist of subsystems made by different vendors • With

Problem • Many industrial systems consist of subsystems made by different vendors • With increasing digitalization, operation and optimization of the system depends on the ability of the subsystems to share data with each other • Unfortunately, sharing data across company borders is beset by both business technical bottlenecks, e. g. – Keeping track of cross-licensing models and other commercial conditions of shared data – Managing and monitoring ownership and IP rights to data, including data derived from other data through modelling and analysis

Potential (partial) solution • Split the life-cycle of shared data in elementary transactions –

Potential (partial) solution • Split the life-cycle of shared data in elementary transactions – Creation of data X 1 by agent a – Grant access of agent a’s data X 1 to agent b, creating a new data X 2 with some specific licencing conditions analogous to Creative Commons licenses – Manipulate input data X 1, …, Xn to create output data Y 1, …, Yn with algorithm Z • E. g. , split data in several parts (partial view), combine several pieces of data into a compound data (combination), … – Deletion of data X 1 by agent a

Potential (partial) solution • A transaction with inputs X 1, X 2 , …,

Potential (partial) solution • A transaction with inputs X 1, X 2 , …, Xn is considered valid if the agent executing it has sufficient rights to all inputs – Creation of data: always valid, the agent creating the data has full rights to the data created – Deletion of data, always valid, the agent deleting the data has no rights to the data anymore – Granting access to the data: valid if donor’s license allows “sublicensing”, receiver’s rights cannot “exceed” those of donors – Manipulating the data: depends on the specific algorithm • This is the tricky stuff …

Potential (partial) solution

Potential (partial) solution

Assignment • Work out details (transaction models, licences) • Encode the validity rules in

Assignment • Work out details (transaction models, licences) • Encode the validity rules in some convenient format (decision table? ) • Work out how to represent the transactions in a blockchain • Potential extension: a micro-payment scheme to implement a business model embedded in digital contracts

Variants • Very interesting variant – Instead of atomic data, the items manipulated are

Variants • Very interesting variant – Instead of atomic data, the items manipulated are continuous data streams (such as sensor outputs such as temperature readings or audio streams) – This probably requires somewhat different transactions and licenses, use your imagination! • Many other variants can be imagined, e. g. , – Keeping track of how licensed source code is being used in a community of developers – Keeping track on the use of media inputs to create new media outputs in a community (e. g. , indy movie makers reusing snippets of each other’s work) • Feel free to suggest your own variant!