Interoperability and Supply Chain Integration

Lead Researcher
Flávio Soares Silva

Associate Researchers
Ana Cristina Vieira de Melo
Antonio Carlos Oliveira Amorim
Arlindo Flavio da Conceição
Ely Bernardi
Flavio Soares Correa da Silva
Frederick Nazario Moschkowich
Gustavo dos Santos Vieira
Lourenço Alves Pereira Júnior
Maria Cristina Machado Domingues
Olga Satomi Yoshida
Rui Tadashi Yoshino
Valéria Nunes dos Santos
Vladimir Emiliano Moreira Rocha

Lines of Research
Interoperability and Supply Chain Integration

A supply chain is a complex and dynamic system composed of multiple decision-making instances (KORPELA; HALLIKAS; DAHLBERG, 2017). Therefore, information about demand, products, production, and services must be accessible and traceable throughout its entire extent.

Integration in business processes is based on supply chain reference architectures and standards that provide integration of product data across all links. In a supply chain, establishing complex interoperability requires mapping and integrating specific data for various organizations and systems (KORPELA; HALLIKAS; DAHLBERG, 2017). Studies by Hallikas et al. (2019), Long et al. (2019), Lin and Zhang (2020), and Pan et al. (2019) have shown the existence of technological gaps concerning information interoperability in a production chain. Some articles mention the application of blockchain (GÖKALP; ÇOBAN; GÖKALP, 2019; MONDRAGON; MONDRAGON; CORONADO, 2020; SCHEID; RODRIGUES; STILLER, 2019), while others discuss M2M (QUEIROZ; WAMBA, 2020) and interoperability in the supply chain.

As the supply chain theme is cross-cutting, the COVID-19 pandemic has heightened the perspective of a broader, deeper, and faster transformation. Consequently, various enabling technologies (SCHWAB, 2017) such as Artificial Intelligence, IoT, Big Data, Cloud, Cyber-Physical Systems, Blockchain, 3D Printing, Digital Twin, Maintenance, among others, will be present in the interoperability and supply chain context to achieve productivity, soft, digital, and hard skills, and cost reduction for the benefit of social well-being. These enabling technologies are interconnected and are challenging to analyze independently; therefore, they should be addressed in a multidisciplinary manner in the Fourth Industrial Revolution.

The introduction of these new technologies into information systems used in the supply chain creates a new demand for studies on the certification of the quality and integrity of the produced information. For systems involving machine learning, new ways of ensuring system quality are necessary, either through testing or verification of their behaviors (ZHANG et al., 2020), or with new development processes (AMERSHI et al., 2019). In these new technologies, system operation depends on both the internal behavior of the software and the data used in learning, which is periodically updated (KHUVIS et al., 2019). Specification, implementation, or interoperability errors in these systems can propagate throughout the supply chain and require new forms of management (NI; XIAO; LIM, 2020) and techniques to ensure software quality during development. In this project, new methods of testing and verification of information systems involved in the supply chain, which include machine learning, will be explored.

The supporting infrastructure to be created by the Collaborative Platform for the construction of collaboration networks requires the establishment of mechanisms for (1) facilitated exchange of products, information, and services, (2) self-regulation of the completed exchanges to ensure fair distribution and quality assurance of resources, and (3) verification of correctness and quality of results generated by intelligent systems.

The establishment of an ecosystem of innovative companies, based on Artificial Intelligence, requires the development of a platform for mediation and quality assurance of products, information, and services exchanged among the companies in the formation of supply chains and collaborative networks.

Within the scope of the Collaborative Platform for AI, this platform will be developed with a focus on three pillars:

  1. Production of an open repository of documents containing specifications of open standards for data interoperability in supply chains. The specifications should include interoperability between systems (machine-to-machine or M2M). These standards will be collaboratively built by the network of innovative companies attracted to the ecosystem nurtured by this project, in alignment with existing standards and practices (BARYANNIS et al., 2019), following processes mediated by technicians from IPT. IPT will be the administrator and curator of the proposed repository.
  2. Production of infrastructure for self-regulated mediation of interactions, based on the concepts of Capability Networks (CORREA DA SILVA et al., 2002), Interaction Protocols (CORREA DA SILVA; ROBERTSON; VASCONCELOS, 2015), and Trust Networks based on Blockchain (AZZI; CHAMOUN; SOKHN, 2019; CONCEIÇÃO et al., 2018; SULTANA; LOCORO; DA SILVA, 2018), enabling reliability assurance in collaborative networks through decentralized mechanisms supported by complex network concepts and behavioral game theory.
  3. Production of technical mechanisms for validation and quality assurance of machine learning techniques and systems, based on empirical software engineering techniques on one hand, and formal verification techniques on the other hand.

The three pillars present significant challenges:

  1. The repository of standard specifications requires active consultations with the community interested in these standards to identify relevant requirements, as well as ongoing research to maintain alignment of the proposed standards with international practices and regulatory aspects, both internationally and locally. This work requires specialized and experienced personnel to ensure that the repository has real practical value.
  2. The infrastructure for mediating interactions will be innovative in nature and will require the development of new techniques to address specific problems related to supply chain logistics in the industry and the integration of production ecosystems based on intelligent systems.
  3. Finally, the mechanisms for ensuring the quality of machine learning techniques and methods are positioned at the cutting edge of artificial intelligence research. In addition to contributing to the ecosystem, these mechanisms will constitute research results with high potential for impact.
  4. Development of methods for data and monitoring system management, aiming to reduce risks to supply chain interoperability.
  5. Development of concepts and means to ensure supply chain interoperability with traceability of data and information reliability in an ecosystem that encompasses various technological pillars already used by the startup and research community, such as computer vision, big data analytics, text analysis, voice recognition, deep learning, and machine learning.

In addition to the scientific impact mentioned in the previous paragraphs, it is expected that the results of this work will effectively have the following consequences:

  • Reliability in real-time monitoring of product origin and destination;
  • Increased trust between producers and consumers, improving the international visibility of domestic producers;
  • Reduction in administrative and bureaucratic costs;
  • Reduction or elimination of fraud and defective product production;
  • Minimized product recalls in real-time;
  • Decreased or eliminated cost of regulation and standardization for regulatory agencies;
  • Lower maintenance costs for companies' asset parks, with increased risk and error tolerance;
  • • Technological support for regulatory agencies.