About this solution
Objectives
The programme develops practical skills in artificial intelligence and machine vision, combining theory and practice: from data acquisition to model design, training, testing and evaluation.
- Fundamentals of AI and ML
- Supervised, unsupervised and reinforcement learning.
- Classification and regression tasks; bias/variance concepts.
- Data and representation
- Complexity, variance, heterogeneous objects and dataset quality.
- Creating/labelling data, train/validation/test split.
- Neural networks and CNNs
- Neurons, weights, activation functions, input/hidden/output layers.
- Convolutions, filters, feature maps and heatmaps.
- Model generalisation
- Overfitting and underfitting; augmentation and regularisation.
- Evaluation metrics and basic interpretability.
- Instruments and integration
- Vision systems with CMOS sensors, autofocus and integrated lighting.
- Embedded AI systems, web/mobile app interfaces and industrial protocols (e.g. PROFINET).
- Immersive learning
- AR environments with Tec2SKILL® ExploreAI: quizzes, animations and hands-on activities.
Aims
The lab consolidates AI concepts by turning them into practical skills through simulations and guided exercises, fostering cooperation, critical thinking and interdisciplinarity across computing, automation, electronics and data science.
- From theory to practice: designing and training neural networks on real vision cases.
- Collaborative teaching: multiple workstations, shared documentation and self-assessment tools.
- Technical and ethical awareness: analysis of the implications and limits of AI applications.
- CPS integration: examples of AI use in cyber-physical stations with data exchange.
- Continuity of learning: structured materials for individual study (manuals, quizzes, apps) and guided lessons.
Conclusion
By the end of the programme, students will be able to:
- Collect and label data from CMOS cameras and prepare it for training.
- Design, train and test neural networks (including CNNs) for classification/regression tasks.
- Evaluate models with appropriate metrics, detect over/underfitting and apply improvement techniques.
- Integrate models on embedded platforms with web/app interfaces and industrial protocols (e.g. PROFINET).
- Use AR with Tec2SKILL® ExploreAI to reinforce learning and check understanding of concepts.
The lab offers an up-to-date, inclusive ecosystem that combines specialist hardware, interactive software and innovative methods for active, collaborative learning that is relevant to professional settings.
Other solutions for:
