Titulo Estágio
Intelligent systems for bill of materials management
Áreas de especialidade
Sistemas Inteligentes
Local do Estágio
Trabalho Remoto (Sede da empresa em Coimbra) & DEI
Enquadramento
Takinobori is a small enterprise with its origins in Coimbra developing Koi [1]. Our focus is on risk management through informed business decisions and option handling.
Ongoing efforts to modernise manufacturing processes are very much aligned with our goals and we're constantly trying to follow the latest developments. One such development is proposed by TerminusDB [2], which addresses the problems of collaboratively building and curating knowledge graphs with native revision control.
In classic bill-of-materials applications rules for the relationships between components can be thought of as grammar rules, as already shown by Clocksin and Mellish [3]. These can be expressed using a definite clause grammar (DCG) as provided by Prolog. However, it is not clear how the potential of logic or intelligent systems can be fully exploited to optimise BOMs, independently. Their dynamic nature has a direct impact on companies and their resilience. Moreover, we need to understand how these systems can be integrated with modern tools, which are typically based in Python and other languages.
We already maintain a reliable and comprehensive system for enterprise resource management and manufacturing, based on FreeBSD and Python, but further research is required.
You will be part of a small but knowledgeable team of professionals, and will have access to real-world challenges, both technical and organisational. Weekly meetings will be scheduled and support given throughout the entire project.
We are looking for someone interested in learning more about intelligent systems, logic, and problem solving in general. Experience in BSDs or Linux will be an advantage.
[1]
Objetivo
The overarching goal of this project is to study state-of-the-art solutions for the management of resources within an enterprise through the use of logic, unification, DCGs, knowledge-based systems, case-based systems, and knowledge graphs. BOM management and optimisation will benefit from artificial intelligence techniques, or expert systems, in order to guarantee an efficient process. Ultimately, this should lead to a stable, well-defined and documented system that increases the resiliency of the entire enterprise and allows new insights over existing business processes and their dependencies.
The outcomes to be achieved include:
* familiarisation with enterprise resource management principles (e.g. manufacturing orders, bill of materials);
* a theoretical analysis and comparison of state-of-the-art solutions (e.g. Prolog and TerminusDB);
* definition of requirements and design of solution for BOM management and optimisation;
* implementation and evaluation in a real testbed.
These outcomes shall be part of the final thesis and, if meaningful developments are made, part of a publication or of a public code-repository.
In addition to the practical objectives, we see the opportunity to strengthen bridges between academia and industry in very important area for the future and heavily emphasised by the European Union: resiliency and digitalisation of enterprises. Finally, we expect to contribute to the training of future professionals in this field.
Plano de Trabalhos - Semestre 1
During the first semester the project shall include:
* A preliminary study of enterprise resource management systems, focusing on Odoo [1 week];
* A preliminary study of knowledge graph systems, focusing on TerminusDB [1 week];
* A literature search of similar approaches [3 weeks];
* A theoretical comparison study of found solutions [1 week];
* Definition of research questions and requirements [3 weeks];
* Choice of one candidate solution and preliminary testing [4 weeks];
* Intermediate Report: documentation of preliminary findings, writing of the state of the art and planning for second semester [4 weeks].
It is important to note that this work will be closely aligned with Takinobori's own systems and developments. This means that the search space on the topic will be contained, with focus on Odoo, Prolog and TerminusDB, but still open to new discoveries. The goal is provide you with initial documentation and materials while also allowing for an independent and critical search/research process.
Plano de Trabalhos - Semestre 2
During the second semester the project shall include:
* Revision of the state of the art, research questions and specification of methodology [2 weeks];
* Refinement of the minimum viable proof-of-concept solution [2 weeks initially and in parallel with other activities throughout the semester];
* Deployment and implementation of the designed solution [6 weeks];
* Validation, evaluation and additional refinements [4 weeks];
* Documentation (writing of thesis) [4 weeks].
As with the work in the first semester, you will be given autonomy for conducting your own research. Nonetheless, Takinobori will accompany the research process until its conclusion, providing input and guidance whenever deemed necessary.
Condições
You will be working as part of a team at Takinobori and will be included in all relevant activities in the company. In addition, the thesis will be conducted in alignment with a member of the ECOS, CISUC group.
Work at Takinobori is conducted remotely, therefore you are expected to have sufficient computation resources (i.e. thin client) to connect to our server/cloud resources. However, we will guarantee access to servers/cloud resources for conducting all the necessary experiments and evaluations.
Depending on the quality of your developed work and overall engagement during the first semester, there is the possibility of remuneration for the duration of the second semester. Furthermore, the best candidates may expect a job offer at Takinobori.
Observações
Supervisor at DEI: João Correia
Orientador
João Pedro Afonso Oliveira de Almeida
education@takinobori.com 📩