1 of 24

Hvordan kan KI-revolusjonen forandre måten vi jobber med og utvikler maskin- og energiteknologi?

Andrei Lobov

e-mail: andrei.lobov@ntnu.no

phone: +47 45 913 455

room: 213

web: http://www.lobov.biz/academia

�Slides:> http://www.lobov.biz/academia/250904

2 of 24

Outline

  • Computer Aided Design : Getting computers to work for engineers
  • Digital threads in modern manufacturing enterprises
  • Towards AI-driven automation of digital threads

2

3 of 24

In the beginning the was a CAD

Dataassistert konstruksjon

3

4 of 24

Computer-aided Design (CAD) - Background

  • CAD used to model product geometries.
  • May record also some non-geometric design specs.
  • Design model can serve as the interface between design and manufacturing.
  • CAD is critical for Computer-aided manufacturing (CAM).
  • Development of CAD began in 1960s ...

4

5 of 24

So, one can design a product (CAD)

5

6 of 24

Computer-aided Manufacturing (CAM)

Analysing manufacturing of workpieces and operations of production plants.

6

7 of 24

One can design a process/production system (CAM)

7

8 of 24

One can continuously integrate product and manufacturing system design

Check the properties of the product / part.�Check the performance of a manufacturing system.�Design for ‘X’ (assembly, quality, resourcefulness, ...)�...

8

9 of 24

One can reuse different models at runtime

9

Process model

Product model

Resources model

NEEDS

CAPABILITIES

MATCH?

If not, how to make it possible?

10 of 24

Digital threads in modern manufacturing enterprises

“En digital tråd er flyten av informasjon om et produkts ytelse og bruk, fra design til produksjon, salg, bruk og avhending eller resirkulering.”

10

11 of 24

Digital thread

11

source: J. David, E. Järvenpää and A. Lobov, "Digital Threads via Knowledge-Based Engineering Systems," 2021 30th Conference of Open Innovations Association FRUCT, Oulu, Finland, 2021, pp. 42-51, doi: 10.23919/FRUCT53335.2021.9599986.

12 of 24

Digital Thread : Implementation (1)

12

source: Joe David, Eric Coatanéa, Andrei Lobov, Deploying OWL ontologies for semantic mediation of mixed-reality interactions for human–robot collaborative assembly, Journal of Manufacturing Systems, Volume 70, 2023, Pages 359-381, ISSN 0278-6125, https://doi.org/10.1016/j.jmsy.2023.07.013

13 of 24

Digital Thread : Implementation (2)

13

source: Joe David, Eric Coatanéa, Andrei Lobov, Deploying OWL ontologies for semantic mediation of mixed-reality interactions for human–robot collaborative assembly, Journal of Manufacturing Systems, Volume 70, 2023, Pages 359-381, ISSN 0278-6125, https://doi.org/10.1016/j.jmsy.2023.07.013

14 of 24

Towards AI-driven automation of digital threads

KI å hjelpe oss til å automatisere digital tråd, men man trenger fortsatt forskjellige språk i den digitale tråden.

14

15 of 24

Digital transformation through semantic modelling

15

16 of 24

ChatGPT “routing” a pipe?

16

17 of 24

Digital transformation through semantic modelling and Business Process Model and Notation (BPMN)

  • To document business processes.
  • To improve communication between business analysts, developers, and stakeholders.
  • To support process automation by dynamically composing and executing the models.

17

Generating BPMN with LLM:

18 of 24

Digital transformation through semantic modelling (3)

18

19 of 24

Takeaways

  • There are standards and languages (representation and programming) available to allow data, information, knowledge representation and processing for various application domains.
  • Also, learning or retrieving knowledge without it being explicitly programmed/hardwired has already some successful ad hoc solutions.
  • So, AI to enable forming and executing “engineering recipes” automatically.

19

20 of 24

Questions?

Andrei Lobov

andrei.lobov@ntnu.no

+47 45 913 455

20

21 of 24

Andrei Lobov

Research mission:I am working on developing a framework for efficient building of knowledge-based �products and knowledge-driven systems, where the knowledge is represented in �machine-processable format thus allowing automated reasoning. I am searching for �and developing an “ideal” language for knowledge representation along with �engineering tools to manage products and smart systems design via their lifecycles �with the help of the tools and techniques originated from the field of Artificial �Intelligence. My research interests include efficient integration of multi-disciplinary �engineering knowledge, its proactive use and support by digital tools.

21

22 of 24

5 schools of thought :�Encoding the rules

  • Symbolists
  • Connectionists
  • Evolutionaries
  • Statisticians
  • Interactionists

22

image source: https://www.w3.org/community/kiss/

23 of 24

Boundaries for evolution / design automation to keep in mind?

“An aircraft optimization starting with a biplane could evolve into a monoplane with a low, mid, or high wing; but a biplane will not arise from a configuration initialized as monoplane.” �[source: Sobieszczanski-Sobieski, Morris and van Tooren, �“Multidisciplinary Design Optimization Supported by Knowledge �Based Engineering”, Wiley, ISBN 9781118492123, 2015]

23

24 of 24

24

�N.V. Bugaev (1837-1903)

Matematician

Teacher for EgorovKolmogorov, Luzin...

image source: https://en.wikipedia.org/wiki/Nikolai_Bugaev

Since no one knows what is an ‘intelligence’, we cannot speak about intelligence in animals, therefore I declare this meeting closed.

The questions of modelling the artificial intelligence are difficult due to absence of ‘clear’ definition of a term ‘intelligence’.