Малайзийская компания Gamuda внедряет GenAI для оптимизации процессов строительства

Malaysian construction and engineering company Gamuda Berhad (Gamuda) is integrating generative AI (GenAI) capabilities into operations to transform traditional construction practices, enhance operational efficiency and improve decision-making.

It will make enterprise-grade GenAI accessible to employees, and help efficiently deliver engineering, construction, and public infrastructure projects in the region.

The company delivers projects in Malaysia, Singapore, Vietnam, Taiwan and Australia, with 4,200 employees.

Gamuda Berhad’s group chief digital officer, John Lim Ji Xiong, said through the Gamuda Innovation Hub, the company is adopting a digital and data-driven approach for construction while upskilling talent.

With Google Cloud as the preferred cloud provider, it becomes easy for the workforce to build GenAI tools to address challenges in work with privacy and security, he added.

Gamuda has developed the Gamuda Digital Operating System (GDOS) – a set of tools consolidating project data on a unified data cloud platform.

This system enables real-time project workflow visibility for enhanced decision-making.

In 2023, Gamuda migrated systems including Autodesk Construction Cloud and SAP S4/HANA to Google Cloud with support from Cloudspace.

This unified data cloud empowers teams including design, engineering, finance, supply chain, and field operations, with integrated and real-time project insights.

Lim said moving compute workloads, including SAP S4/HANA to Google Cloud, allowed the company to harmonise compute. Data in a single cloud platform, thereby enabling the team to focus on driving value-creating use cases and shifting the focus away from managing infrastructure.

Google Cloud’s BigQuery data warehouse and workload-optimised infrastructure helps in cost savings, is easy to use, and helps derive value from data footprint, he added.

The company has implemented Google Cloud’s Security Command Center Premium platform, leveraging machine learning for threat detection, prevention, and regulatory compliance maintenance.

This helps safeguard enterprise data and core digital systems.

GenAI for tunnelling operations

Gamuda uses Google Cloud’s Gemini models on the Vertex AI platform to build and integrate a GenAI-powered conversational agent into its cloud-based Tunnel Insight platform.

This is done with support from CloudMile – an AI and cloud service provider.

Tunnel Insights ingests, presents, and analyses sensor data from Gamuda’s in-house developed autonomous tunnel-boring machines (A-TBMs), which are used to automate repetitive tasks and improve tunnelling operations.

According to Gamuda, these A-TBMs are utilised in the Defu and West Coast Mass Rapid Transit stations in Singapore and the Sydney Metro West-Western Tunnelling Package in Australia construction projects.

The volume of data generated presents challenges in extracting timely insights for responding to geological changes or maintenance needs.

With GenAI-powered conversational agents, staff can extract information from machine documentation and interpret performance data in natural language.

This integration improves operational efficiency and enhances safety in tunnelling projects.

GenAI to help tender proposals

The company is using Vertex AI Search and Conversation to build generative search and chat applications for market intelligence, design, and technical teams.

It will help synthesise research documentation into concise summaries and extract insights from past projects for new tender proposals.

Gamuda aims to empower employees to develop customised GenAI applications tailored to specific roles through an internal marketplace called BotUnify.

Employees have control over the data sources accessed by these applications, ensuring the integration of enterprise data and the implementation of features including source citations to enhance response quality.

Google Cloud Malaysia’s country manager, Patrick Wee, said Gamuda is merging data streams to solve real-world challenges while saving time and costs.

It is supporting Gamuda’s GenAI-powered advancements across autonomous drilling, Building Information Modeling, and augmented reality, Wee added.