From Mechanical Engineering to AI: Ranitabh Mallick’s Erasmus+ Journey
How an international learning experience in Romania connected a Gisma student’s engineering background with artificial intelligence, sustainability and predictive maintenance.
What happens when an engineering background meets artificial intelligence and international collaboration? For Gisma student Ranitabh Mallick, an Erasmus+ programme in Romania provided an opportunity to explore this question through a practical sustainability project, teamwork and new technical approaches.
Ranitabh is pursuing an M.Sc. in Data Science, AI and Digital Business at Gisma University of Applied Sciences. Having previously studied Mechanical Engineering and worked in operations, he wanted to develop his technical skills and explore how data science could improve processes and support better decisions. His experience in Romania became one step in this journey, influencing the projects and research he pursued afterwards.
Taking learning beyond the classroom
Ranitabh was nominated by Gisma to participate in the Erasmus+ Blended Intensive Programme at Babeș-Bolyai University in Cluj-Napoca, Romania. The programme focused on AI and sustainability, bringing together participants from different academic and cultural backgrounds.
For Ranitabh, it was an opportunity to represent Gisma, exchange ideas and apply his knowledge to a practical challenge. Working towards a shared project also highlighted the importance of collaboration and clear communication, particularly when presenting technical concepts to an audience.
The experience helped him recognise that developing a solution involves more than making it work. Explaining its purpose, demonstrating its value and communicating the results are equally important.
Developing the Carbon Footprint AI Assistant
During the programme, Ranitabh and his team developed the Carbon Footprint AI Assistant, a prototype designed to make transport-related emissions easier to understand.
The project calculated estimated emissions using travel information and used AI to explain the results and suggest practical changes. The team used n8n to connect the workflow steps and Mistral through Ollama to generate recommendations locally.
As the AI Recommendation, Explainability and Quality Assurance Lead, Ranitabh focused on making recommendations understandable and ensuring they were supported by the underlying calculations.
An important technical principle was separating numerical calculations from AI-generated explanations. The calculations provided the basis for the results, while the AI helped communicate their meaning. This approach highlighted the importance of testing, data reliability and clearly defined responsibilities within an AI workflow.
Ranitabh achieved 97% and received recognition for his final presentation and representation. The result marked an important milestone in his transition from Mechanical Engineering and operations towards data science and AI.
Applying the learning to practical projects
Following the programme, Ranitabh continued exploring how automation and AI could support decision-making through independent projects.
DelaySense Berlin combines live Berlin public-transport departure information with rule-based delay-risk scoring and local AI explanations in an n8n workflow. The prototype supports decisions such as leaving earlier, waiting or considering an alternative. Ranitabh has documented the project, its testing and its workflow on GitHub.
His second project, AutoPass Sentinel, focuses on battery-passport readiness and data quality. Using n8n and deterministic validation rules, the prototype checks structured information, identifies missing evidence and produces outputs that can be reviewed.
Although the projects address different challenges, they share a common approach: collect information, validate it, apply defined decision logic and make the results understandable. These projects have allowed Ranitabh to explore how AI and automation can be applied while keeping the underlying processes transparent and reviewable.
Connecting engineering with predictive maintenance
Ranitabh has also applied his interests to his master’s thesis, titled “An Uncertainty-Aware, Explainable and Human-Reviewed Framework for Remaining-Useful-Life-Based Maintenance Prioritisation.”
Using the NASA C-MAPSS FD001 turbofan dataset, his research examines machine-learning approaches for estimating the remaining useful life of equipment and supporting maintenance prioritisation.
He compared machine-learning approaches, evaluated a selected Random Forest model, used conformal prediction to examine uncertainty and applied SHAP to help explain model predictions.
Alongside the modelling, he developed a separate n8n human-review prototype. It uses a fixed illustrative input, applies maintenance-priority rules, allows a local AI model to provide a bounded explanation and records whether a human reviewer approves, rejects or modifies a proposal.
The research brings together his engineering background and developing expertise in data science. It also reflects his interest in understanding how predictive insights can be translated into operational decisions while accounting for uncertainty and human judgement.
Lessons from an international experience
The Erasmus+ programme helped Ranitabh consider AI as part of a broader decision-making process. Beyond developing a model, this means understanding where the data comes from, how it is validated, how results are communicated and when human review is appropriate.
Working with people from different backgrounds also reinforced the importance of explaining ideas clearly and approaching shared challenges from different perspectives.
For students considering similar opportunities, Ranitabh encourages them to apply, ask questions, contribute to team projects and look for ways to develop their learning further afterwards.
His experience demonstrates how an existing academic or professional background can provide a foundation for exploring a new field. From the Carbon Footprint AI Assistant in Romania to DelaySense Berlin, AutoPass Sentinel and his predictive-maintenance thesis, he has continued to connect his studies with practical applications.
For Ranitabh, the Erasmus+ experience was one part of a wider learning journey, offering ideas and approaches that he has continued to explore through independent projects and academic research.