The COVID19 outbreak has brought the world to a standstill and has disrupted the aviation industry like no other crisis in modern times, according to Oliver Wyman’s Airline Economic Analysis. Unprecedented collaboration and greater data sharing in the aviation data value chain are identified by Deloitte as key factors to ensure that aviation emerges from the pandemic more responsive, more agile and more sustainable than ever before. Therefore, despite the circumstances, data sharing remains at the top of the agenda in aviation with IATA acknowledging it as a key aspect for the proliferation of data science in aviation and specifically stating that “an open culture towards data pays dividends when being part of a trusted and governed sharing ecosystem”.
Sharing should not be only about the data
Although the majority of data marketplaces typically focuses on the exchange of data, sharing should consider not only raw data, but also derivative data (as results of an analysis over raw data) and non-data assets (in terms of analytics models and workflows that use the raw data). The ICARUS Platform supports aviation data asset brokerage in terms of both raw data and non-data assets (i.e. applications). However, sharing an application poses a number of implications and challenges related to the IPR: an application consists of a trained model and the raw dataset(s) used for training and/or testing purposes, that may belong to one or multiple data asset providers, thus the data asset sharing contract needs to be signed among all involved stakeholders, addressing all legal aspects related to intellectual property sharing. In order to homogenise the brokerage process under a common framework to increase efficiency and the demand for customised rights, terms and conditions and negotiation mechanisms in the complex landscape of data-enabled assets’ IPR, ICARUS allows sharing an application depending on the privacy level, the license metadata and the ownership of all its “ingredients” (that should be common under the same stakeholder).
License compatibility for hybrid, smart contracts is hard to automate
ICARUS foresees both the creation and brokerage of assets that constitute combinations of other underlying assets and/or derivative work of previously existing assets (with the help of Data Preparation and the Analytics Workbench functionalities). This places license compatibility analysis at the core of the challenges that need to be addressed in regard to asset sharing that can be automatically checked up to an extent (based on the raw dataset’s license metadata). Nevertheless, it cannot be provided as a fully automated process at this stage, as due to the dual nature of the smart contracts (containing terms and conditions in text, not just in the structured metadata) it would require natural language processing on legal documents, which is not in the scope of the ICARUS platform.
Negotiation is key to reach consensus for data asset sharing
Data asset sharing is not about the data asset provider imposing the terms and conditions of the transaction, but both the data asset provider and the data asset consumer need to reach consensus on terms and conditions that are mutually satisfactory and beneficial. To this end, the negotiation functionality that allows keeping a record and reviewing step-by-step the changes introduced by the other party is instrumental in the ICARUS Platform (even though writing all intermediate versions of a smart contract in the blockchain is costly from a performance perspective and requires the involved parties to wait for the transaction to be mined).
More details on the ICARUS assets brokerage framework can be found in our previous blog post.
Blog post authored by Suite5.