Thematic Panel on Construction concludes that digitalisation is advancing faster in research papers than on construction sites

The thematic panel «BIM, Artificial Intelligence and Sustainable Management of Construction Projects» was one of the activities held as part of the 30th International Congress on Project Management and Engineering (CIDIP / ICPME 2026).

Participants:

• Carlos J. Pampliega García (Vice President of the PMI Madrid Chapter)
• Juan Manuel David Rodríguez (PhD candidate in Water and Environmental Engineering, Universitat Politècnica de València)
• Kadhum Atabi (Civil Engineer and Consultant, PhD, Universitat Jaume I)
• Susana Torno Lougedo (Associate Professor at the University of Oviedo and specialist in computational fluid dynamics applied to mining)
• Rosa Ramírez de Arellano Marrero (Predoctoral Researcher at the University of Seville and recipient of an AEIPRO Best Master’s Thesis Award)

Moderator:

• M. Socorro García Cascales (President of AEIPRO–IPMA Spain)

The session brought together five highly diverse papers, ranging from a global bibliometric review of the PMI-CP™ certification to laboratory testing of bentonite suspensions for pile foundations.

Five different research scales, five countries of origin, and one common underlying question: how can the complexity of construction projects be transformed into data-driven decision-making before problems arise on site?

M. Socorro García Cascales analyses the main conclusions of a session that successfully connected research, technology, and professional practice.

From the discussion, four key ideas emerged which, in my view, define where the construction sector truly stands today in relation to digitalisation:

1. The gap between professional practice and academia remains the main challenge

Pampliega García’s bibliometric review provided the ideal starting point. His analysis of the PMI-CP™ certification shows that professional practice devotes 50% of its content to contract management, while academic literature is already exploring artificial intelligence, blockchain, and smart contracts to prevent disputes before they occur.

The same tension reappeared, almost like a mirror image, during Kadhum Atabi’s presentation on construction and demolition waste legislation. While the European Union has established a binding recovery target of 70%, Iraq manages construction waste solely through individual project plans, without a specific legal framework to support them.

In both cases, regulation lags behind what science—and even informal practice, such as the spontaneous recycling of steel and aluminium highlighted by Atabi—is already achieving in the field.

The panel reached a clear consensus: it is not that regulation «fails», but rather that it serves a different purpose from research, and both must learn to interact much more rapidly.

2. Scale as a challenge in translating knowledge

Rarely does a panel bring together such contrasting research scales. On one hand, Pampliega analysed global patterns based on thousands of Scopus-indexed publications. On the other, Susana Torno Lougedo presented laboratory findings on the thickness, measured in millimetres, of the filter cake formed by bentonite slurry on the wall of a pile foundation—a remarkably precise result showing that contamination with just 1.96% sand doubles the filter cake thickness, while the use of a filtration reducer completely reverses the effect.

Between these two extremes, Juan Manuel David Rodríguez demonstrated how three artificial intelligence techniques (MLP neural networks, CatBoost, and SVR-RBF) predict cavitation phenomena in hydraulic valves with an accuracy exceeding 99%, based on 855 experimental tests.

The challenge raised by the panel is significant: how can such highly specific and valuable laboratory findings be translated into competencies that professional project management standards can genuinely capture and teach?

No definitive answer emerged, but there was a shared understanding that bridging the gap between technical detail and managerial competence remains an unfinished task.

3. Sustainability as the common thread running through all five papers

Although the panel was not explicitly conceived as a sustainability session, sustainability naturally became the common thread linking all five presentations.

Kadhum Atabi addressed it from the perspective of public policy and the circular economy for construction waste.

Rosa Ramírez de Arellano Marrero, through her review of the integration of BIM, Life Cycle Assessment (LCA) and Artificial Intelligence in industrial facilities, demonstrated how algorithms such as Gradient Boosted Trees and Random Forest can already predict environmental impacts with errors ranging from just 0.90% to 17% compared with a full life cycle assessment. However, she also identified a critical gap: nearly all previous studies focus on residential or office buildings rather than industrial facilities with complex pressurised systems.

Even Susana Torno Lougedo’s apparently more technical work on bentonite ultimately contributes to sustainability through improved structural stability and the prevention of costly failures.

Pampliega’s bibliometric study confirmed the same trend: BIM consistently appears linked to sustainable development in recent keyword co-occurrence analyses.

Sustainability, we concluded, is no longer a separate chapter in construction project management; it has become a cross-cutting dimension influencing every decision, from legislation to the millimetre-thick filter cake formed during pile construction.

4. The future: digitalisation is advancing faster in research papers than on construction sites

Perhaps the panel’s most honest—and uncomfortable—conclusion was that there is still a considerable gap between what research is already capable of achieving and what actually happens on construction sites.

Pampliega’s density maps illustrate this clearly: BIM, blockchain and artificial intelligence continue to appear as peripheral topics compared with the traditional core of contract management, risk management and stakeholder management.

Rosa Ramírez de Arellano Marrero pointed out that most BIM-LCA-AI frameworks identified in her review remain academic prototypes that have yet to be validated in real industrial projects.

Both Rosa Ramírez de Arellano Marrero and Susana Torno Lougedo reached the same conclusion from different perspectives: the shortage of reliable, traceable and well-structured data—whether in the form of incomplete manufacturers’ performance curves or BIM models with insufficient levels of development—remains the principal obstacle preventing these technologies from moving from the laboratory to construction practice.

Digitalisation, ultimately, is progressing much faster in scientific research than in everyday industry practice, and closing this gap will undoubtedly become one of the greatest challenges for project management over the coming decade.

Final conclusion

We closed the panel with a shared conviction: BIM and artificial intelligence are not isolated technologies, but rather the emerging response to a structural problem within the construction sector—the long-standing lack of standardised and traceable data throughout every stage of the project lifecycle.

My sincere thanks to all five panellists for a rigorous, honest and, above all, highly valuable discussion that continues to build bridges between research and the professional practice of project management.