in processes

Artificial Intelligence

Artificial Intelligence (AI) is a multidisciplinary research area framed within Computational Sciences.

The objective of AI is to emulate human reasoning, so that machines can learn data flows and automatically take decisions.

Although work in this area dates back over one century, it is now, with the technological advances and information explosion, when it has been possible to implement methods that are successfully applied in industrial contexts.

We are currently living in the Digital Era and practically any automatic device has digital electronics, and consequently, processes and transmits data. This has had a boom effect on the industry. Any modern installation processes and transmits loads of data. Thus, a fantastic context has been created to implement AI methods, and this is the basis of the so-called Industry 4.0 (I4.0).

At Lortek, we focus on AI applications for industrial contexts, the majority of which are encompassed in I4.0. Several technologies are involved: Cybersecurity, BlockChain, Intelligent Control, Computer Vision, IoT, Edge-Fog-Cloud Computing or Machine Learning are essential parts.

All of them are governed by basic principles such as:

  • The Human Being as the main figure.
  • Ethics, legality, and transparency in the AI algorithms.
  • Sustainability and efficiency.
  • Compliance with European regulations.

What we are currently working on

3L: Machine Learning, Deep Learning and Active Learning

Apart from working with existing tools in the state of the art (tensorflow, keras, Caffe, sk-learn, etc.) and the different programming languages, at LORTEK we design specific algorithms for industrial problems. In the information explosion in the current I4.0, we address the problem of dealing with different origins and types of data, their traceability, and the detection of false data. We create ad-hoc algorithms, to generate high-precision and quality prediction models.

An aspect that concerns us is knowledge obsolescence. Thus, we accompany our methods with Self-Learning techniques that guarantee maximum performance both in the industrial installation, and in the following tolerance of context changes (changes in product, parameters, or components).

We apply these methods to: predictive maintenance, product optimization, energy optimization, and fault prediction.

Computer Vision

Digital Image Processing is a specific case where Machine Learning methods can be applied, with the advantage that the images are a body of sorted data. In this context, we apply techniques that adapt to the problem of vision, such as analysis of reflectance, chromatic calibration, and automatic systems to correct changes in lighting.

Regarding computational methods, we continue working with traditional methods such as SVM, RF, ANN, Bayes, etc. But, concerning vision algorithms, we also work with the latest Deep Learning advances (CNN, RNN, GAN, etc.).

We support vision systems with Active Learning software architectures in many cases. This means that expert worker is still a key figure for the quality control. Today’s industry demands the manufacture of increasingly shorter series with an increasingly greater variety of products. Expert workers are the ones, who, with human interfaces/AI, manage this process of adapting the algorithms to new manufacturing contexts.

We essentially apply these methods to quality control systems. We observe the product with cameras, and then we conclude if the part is good or not.

Smart Control

Smart control systems use several artificial intelligence techniques to control complex processes. Processes, which models are very complex, non-linear, uncertain environments, or with control criteria and objectives that change over time. Smart control permits solving control problems that are impossible to address with traditional methods. The four basic techniques that essentially comprise Smart Control are Expert Control, Fuzzy Control, Neuronal Networks, and Genetic Algorithms. At LORTEK, we are working on the implementation of these techniques in complex industrial processes with the main objective of improving quality.

Specific Equipment

Our researchers work with powerful workstations with the capacity to manufacture and execute prediction models. We also have dedicated servers with several Xeon processors and NVIDIA cards.

Publications and downloads

MR Parsha Pahlevannejad, Arnold Herget, Ramón Moreno, Andre Hennecke.
Implementation and Testing of a Modular System Architecture for Generic Hybrid Production Cells in an Industrial Environment.
45st Annual Conference of the IEEE Industrial Electronics Society.
Ander Muniategui, Aitor García de la Yedra, Jon Ander del Barrio, Xabier, R. Moreno.
Mass production quality control of welds based on image processing and deep learning in safety components industry.
Proc. SPIE 11172, Fourteenth International Conference on Quality Control by Artificial Vision, 111720L
A Muniategui, JA del Barrio, XA Vinuesa, M Masenlle, AG de la Yedra, R. Moreno.
One Dimensional Fourier Transform on Deep Learning for Industrial Welding Quality Control.
International Work-Conference on Artificial Neural Networks, 174-185.
R. Moreno, Juan Carlos Pereira, et al.
Time Series Display for Knowledge Discovery on Selective Laser Melting Machines.
20th International Conference on Intelligent Data Engineering and Automated Learning.
R. Moreno, E. Gorostegui-Colinas, PL de Uralde, A Muniategui.
Towards Automatic Crack Detection by Deep Learning and Active Thermography.
International Work-Conference on Artificial Neural Networks, 151-162.


Challenges to be faced in the coming years:

The challenges of AI in Industry 4.0 are greater in the context of information explosion. New agents have appeared in the industrial scenario such as: Cybersecurity, IoT, Blockchain, Decision-making help systems. Algorithms can take decisions in some countries and impact the executions of industrial systems located in other countries. Moreover, expert workers provide the inputs of some algorithms. These aspects are becoming increasingly important, requiring us to judge aesthetic and legal aspects of AI.
Apart from the legal and ethical limits of algorithms, at Lortek we address machine learning paradigms that adapt to current industrial needs: systems able to learn on their own from zero (Reinforcement Learning, Self-Learning). In digital image processing, we are working to create Learning without Forgetting systems. Both techniques together have the important objective of creating computational models that compile industrial knowledge, and adapt to the contextual changes required by today’s manufacturing systems: customer requirement oriented product, with increasingly shorter series.
The application of digital solutions to industry has sped up over the last few years. In this sense, the advances in the field of Automatic Control, Artificial Intelligence, sensoring, signal processing, actuators, etc. provide new tools for system control. However, their application is still a challenge in certain complex industrial environments (such as arc welding processes, laser processes, etc.), as a more robust technology with a real-time response is demanded.