Dennis Wong (Class of 2015) is currently a Project Geologist at Arup, holding the professional qualification of Chartered Geologist (CGeol). Through his diverse participation in numerous infrastructure and underground excavation projects, Dennis has honed his skills in ground modelling, ground characterisation, and ground investigation. He has also extensive experience in slope engineering and natural terrain hazard studies. As the Arup’s Regional Rock Mechanics Skill Leader, Dennis is dedicated to bringing overseas skillsets to the region and bridging the skill gaps between engineering geology and tunnelling.

He also actively involved in the Geological Society of London Hong Kong Regional Group, acting as the member of the main committee and the chairperson of the Young Fellow Group between 2019 and 2024.

In addition to his technical pursuits, Dennis has taken up teaching roles at CUHK and HKU Space, committed to nurturing future generations of professionals by disseminating practical knowledge and imparting industrial insights.

What critical mindsets and attitudes do you believe every geologist should possess? 

Ground data can be highly unstructured, diverse, and heterogeneous. Geologists need to approach their work with integrative thinking and strategic planning to piece together the full story from scattered clues, while also understanding which pieces of the puzzle to address and prioritise first.

Developing conceptual models is a fundamental skill that all geologists should master. Equally important is the ability to translate these conceptual models, largely based on individual knowledge and experience, into observational models by incorporating interpretations of a substantial amount of ground data. This transformation process can be influenced by confirmation bias and the formation of information echo chambers, which can lead us astray from the truth.

I think geologists should adopt a scientist's thinking approach, which is advocated in Adam Grant's book "Think Again", to logical deduction by rigorously testing hypotheses through multiple rounds of questioning, in particular when data become available at different phases. This necessitates geologists to grasp their knowledge from first principles, challenging and refining their interpretations from the ground up.

Can you provide examples of projects where you successfully integrated geology into engineering solutions? 

I have been actively involved in the MTR’s Siu Ho Wan Depot Property Development Consultancy Agreement No. C1701 project since its inception in 2021,  involving the provision of detailed design services. The proposed development at Siu Ho Wan Depot encompasses a range of facilities, including rail facilities, road works, schools, shopping malls, recreational spaces, residential buildings, and auxiliary structures. Situated on a piece of reclaimed land in North Lantau, the project poses challenges due to complex regional geology and difficult ground conditions caused by the reclamation setting.

My role in the project involves detailed geological assessments, which include reviewing natural terrain hazards, studying previous construction drawings, interpreting aerial photographs, characterising ground and groundwater conditions, interpreting laboratory test results, and identifying key geological and geotechnical constraints. I have also assessed the adequacy of ground investigation data, identified information gaps, and proposed additional ground investigation works. This allows me to proactively communicate with clients and coordinate with the interdisciplinary team. Throughout this process, I have honed my 3D ground modelling skills and gained insights into ground risk quantification.

What would be your advice to geologists to embrace the AI era? 

It is apparent that the emergence of cutting-edge AI-enabled tools and Large Language Model-based (LLM-based) chatbots is assisting with automated data processing, key data extraction, and preliminary knowledge asset interrogation. To leverage these tools properly and effectively, it is necessary for geologists to assess and fact-check the outcomes with their fundamental knowledge.

Apart from IQ (intelligence quotient) and EQ (emotional quotient), YQ (why quotient) is gaining significance in a VUCA (volatile, uncertain, complex, and ambiguous) world shaped by AI and constant changes. Data are more accessible than never before, meaning the curiosity to explore beyond the obvious as well as the ability to ask the right questions and connect knowledge domains are becoming more powerful.

Staying curious, thinking critically, and challenging assumptions can allow us to stay competitive in the AI era.

What are your perspectives on the future of engineering geology? 

Achieving Net Zero emissions and the United Nations' Sustainable Development Goals necessitates significant involvement from geologists and other ground specialists in extracting essential resources for energy transition and constructing infrastructures for sustainable and efficient energy generation, storage, and delivery. Examples include deep excavations for nuclear waste disposal and geological hydrogen storage. It is disheartening to witness a decline in the number of students pursuing earth sciences or geology. This is also the primary drive for me to take up teaching roles in various education institutions in the hope of inspiring younger generations to the interesting components of geology.

Moreover, understanding the ground is crucial for all structures built on it, as inherent ground risks exist. Geologists should have the best understanding of the ground. With our expertise, design of the structures can be better optimised to achieve low-carbon solutions.
In the future, AI may assist in data processing, pattern recognition, and providing preliminary interpretations. Nevertheless, it is unlikely that AI can totally replace skilled geologists in the foreseeable future, as comprehensive interpretations involve a scientific mindset and a geological conceptual model. This always requires a smarter brain behind the AI to review the outputs.