Page 11 - Unlocking the potential of trust-based AI for city science and smarter cities
P. 11
In a smart city, user-related information works as oil to fuel the state of art applications and
services. Consumers, who use these services, provide personal information to service providers,
intentionally or unintentionally and often without considering their trustworthiness. However, this
personal information often reveals one’s identity and may lead users to face unexpected outcomes,
ranging from uninvited advertisements to identity theft. To regulate such issues, this approach
investigates state of the art data governance techniques that are built on trust, blockchain, and the
distributed AI concepts.
As the aim of a smart city is to make decisions about its data in a more trustworthy manner and
meeting the essential KPIs, the proposed trust-based data management solutions has enormous
potential to securely process and handle data of any service providers or customer.
Further combination of blockchain and IoT will facilitate the sharing of AI services and resources
leading to the creation of a marketplace of services between devices.
For cross-border applications, it can serve as an intermediate broker to handle the negotiations
for particular interaction without any ambiguity or human intervention which outsmart current
techniques based on third party regularity bodies.
With great interest in artificial intelligence around the world today, the approach has the potential
to open up a new chapter in human-machine interaction by giving interoperability to existing AI
technology and combining it with the trust-based data governance concepts.
The proposed Trust-based AI cross-domain microservices across Roof, Fog and Cloud continuum will
not only support the development of real-time applications that address latency and bandwidth
related problems of the current systems but also privacy leakages and security attacks. It will also
support plug and play AI reusable and dynamically composable components that are deployed as
microservices for the development of value-added cross-border use case applications.
2.2 Implementation
Figure 2 illustrates a conceptual system model for data collection, processing and sharing in Smart
Cities from various data sources, including personal data. In a Smart City big data architecture,
collected data are processed and stored in a structured format that can be queried using analytical
tools in an analytical data-store for supporting querying from Smart City agencies. Also, data is
shared with third-party service providers through an open data-store after conducting publishing
control mechanisms.
According to the GDPR [8] legislation in Smart City contexts, citizens (i.e., Data Subjects) authorize
Smart City operators (i.e., Data Controllers) to control their personal data. Data Controllers
determine the purposes for which, and the method in which, personal data is processed by Data
Processors (i.e., Smart City agencies and third-party service providers) - who will be responsible for
processing the data on behalf of the controllers. Therefore, Data Controllers are subject to comply
with requirements and obligations imposed by GDPR when determining personal data usage policies
for Data Processors.
Unlocking the potential of trust-based AI for city science and smarter cities - October 2019 5