What Does it Take for AEC Firms to Successfully Adopt SCALABLE AI?
By Engineering Management Institute
Key Concepts
- Business Transformation
- Scalable AI Adoption
- Process is King
- Data Governance
- Content Management
- SharePoint Backbone
- Microsoft Copilot
- Data and AI Summit
- Personal Productivity (No-code, Low-code AI)
- Pro-code Solutions
- Operations (OpEx) Team
- Business Process Mapping
- ERP Replacement
- Field Data Collection
- Lab Solution
- AI Agent on Data Quality
- Data Architecture
- Curiosity (Core Value)
- Workforce Readiness Certification
- Social-Emotional and Mental Components of Business Readiness
- Energy Consumption of AI
- Data Staging
- Environmental Cost
Introduction to AI Transformation in AEC
The discussion features David Harwood, P.E., Senior Vice President of Business Transformation Information at Terracon, on the AEC AI and Tech Strategy Podcast. The central theme is how data, process, and governance form the bedrock for sustainable AI adoption within the Architecture, Engineering, and Construction (AEC) industry. The conversation covers aligning strategy with data-driven decisions, balancing innovation with governance, and cultivating a workforce culture capable of scaling with AI. The episode also highlights the AECPM Connect event for project managers and Track 3D, an AI-first reality intelligence platform for construction monitoring.
David Harwood's Journey to Business Transformation
David Harwood's career path was non-traditional, starting as a land surveyor in a family business in California, then transitioning to geotechnical engineering. He joined Terracon in Colorado, where he faced an immediate challenge with a depleted department. Through mentorship, he gained a deep understanding of geotechnical engineering and client needs. After earning an MBA from the University of Colorado, Colorado Springs, he pioneered Terracon's first business sector program in transportation, focusing on alternative delivery, design-build, and public-private partnership (P3) projects.
His involvement in senior leadership and strategic planning revealed a critical issue: inconsistent work pricing, which hindered the company's objectives as an employee-owned entity. Harwood initiated a pilot pricing solution in January 2020, which, despite launching just before COVID-19, took a year to build and significantly improved pricing consistency and accountability. This success led to the evolution of his role into business transformation, eventually encompassing all data, AI, governance, and operational excellence (OpEx) solutions. This journey exemplifies how impactful innovation, even from a non-IT background, can lead to the creation of new, critical leadership positions.
Defining Business Transformation and Scalable AI Adoption at Terracon
At Terracon, business transformation, particularly concerning scalable AI adoption, is approached with the understanding that existing business operations must continue without significant disruption. The strategy focuses on identifying foundational elements critical for success. David emphasizes that "process is king, after process is data."
Terracon, with 7,800 employees and nearly 200 offices, historically granted significant autonomy to its offices to tailor services to local markets and clients. This autonomy, while a key to success, created challenges for enterprise-wide data consistency.
The initial steps focused on data:
- Addressing Data Disconnection: Terracon faced a proliferation of disconnected systems, varied data structures, and inconsistent terminology (synonyms and homonyms) across the enterprise.
- Establishing Data Governance: A dedicated data governance team was formed to catalog all existing data, build a comprehensive glossary, and align it with an agreed-upon taxonomy.
- Implementing Content Management: To resolve issues with document filing, retention policies, and email management, a content management solution was built on a SharePoint backbone, leveraging Terracon's existing M365 environment. This transition, launched on January 1, was a "non-event" because employees were already struggling with office-based servers for collaboration and were familiar with Microsoft OneDrive. This provided much-needed structure and content rules.
Leveraging AI for Productivity and Process Improvement
With structured data and content in place, Terracon began exploring AI applications:
- Copilot Pilot Program: An initial pilot of 300 Microsoft Copilot licenses took over a year to distribute because users struggled to find practical applications.
- Data and AI Summit: A turning point was a Data and AI Summit where Microsoft demonstrated Copilot's capabilities to service line leaders, executive teams, and operations leads. This direct demonstration led to an immediate surge in adoption, from 300 to 800 licenses almost overnight. Employees now actively share new use cases, highlighting significant time savings in tasks like crafting emails, writing paragraphs, and locating information.
- Focus on Personal Productivity: Current AI efforts are largely centered on "no-code, low-code" solutions to enhance individual productivity for daily tasks.
- Transition to Pro-code Solutions: The next phase involves building "pro-code" solutions to drive organizational processes. An OpEx team is dedicated to mapping business processes, identifying data elements, systems, time, and labor categories involved. This detailed mapping allows for targeted technology application to achieve measurable savings and ROI.
Process Changes and Digital Transformation Examples
Terracon's digital transformation includes significant process changes:
- ERP Replacement: An ongoing enterprise resource planning (ERP) system replacement incorporates automation, robotic process automation (RPA), and machine learning.
- Field Data Collection and Lab Assignments:
- Previous State: Geotechnical exploration data collection and lab assignments were paper-based, leading to lost information and lack of visibility.
- New Solution: The material and geotech service lines collaborated with a vendor to develop a new field data collection system for geotech and an integrated lab solution.
- AI Application: This standardized process enables the future deployment of an AI agent on data quality. This agent would provide real-time validation of incoming field data against expected ranges or depositional environments, offering immediate feedback for correction.
- Operations-First Mindset: The approach is to identify operational bottlenecks and problems first, then build technology solutions, rather than finding a "shiny tool" and searching for a problem to solve.
- Flexible Data Architecture: Terracon is building a flexible data architecture that can accept content from various new tools and systems. This architecture makes data available through a central data store for consumption by any other solution, fostering innovation from the ground up and avoiding "sorry, no can" responses to new technology proposals.
- Culture of Empowerment: The organization fosters a culture where positive change is driven by those closest to the work, creating opportunities for experimentation and sharing. The rise of low-code/no-code tools empowers individual contributors to develop proof-of-concepts and effectively communicate their ideas.
The Role of Data Governance in AI Readiness
David Harwood emphasizes that "data is that foundational element that allows all of [AI and machine learning] to function." Therefore, robust data governance is critical for ensuring quality and valid data that is "fit for use" and within expected technical ranges.
Challenges and Solutions in Data Governance:
- PM Burden: Project Managers (PMs) often lack time for extensive data entry (e.g., 100 pieces of information for project registration), needing only a project number to start work.
- Inconsistent Data: Issues like varied spellings for the same client ("Walmart" vs. "Wal-Mart") create data inconsistencies.
- AI-Powered Data Gathering: AI can automate the extraction of information (services, clients, addresses) from proposals and contracts, reducing the burden on PMs.
- Shifting Responsibility: Data gathering responsibility can be shifted to departments like marketing, which have a vested interest in accurate client data.
- Reinforcing Governance through Tools: Tools like the pricing solution are designed to enforce data governance. It's the only way to price a project, automatically pulls in metadata, and converts project-specific scope into accounting system requirements, making it easier for users while ensuring data quality. The principle is to automate requirements before asking humans to perform additional tasks. This "operations-first mindset" uses technology to make employees more efficient.
Balancing Experimentation with Governance
The rapid pace of AI tool development (e.g., "everybody's got an agent now") necessitates a balance between innovation and structured governance.
- Prioritization Framework: Terracon evaluates new technologies by understanding the processes they aim to solve, assessing their return on investment (ROI), value to the organization, and potential disruption. This allows for effective prioritization.
- Maturity Assessment: Some advanced technologies might not be suitable for the company's current maturity level. Understanding future capabilities helps shape current implementation steps to prepare for later adoption.
- Long-Term Vision: Terracon maintains a long-term view, considering future technologies like quantum computing (5-10 years out) and ensuring current AI deployments contribute to future advancements. This approach mirrors the incubation period seen in academia and commercial product development (e.g., early OpenAI models vs. ChatGPT). This requires patience and a forward-looking perspective.
Workforce Readiness and Organizational Culture
People are the backbone of any transformation. Terracon's core value of curiosity is central to workforce readiness, encouraging employees to experiment, learn, and understand the "why" behind new approaches.
Broader workforce readiness initiatives include:
- Technology and Data Awareness: Programs, like those in Kansas, focus on certifying workforce readiness, including proficiency in common business tools (e.g., M365) and understanding AI terminology and business applications.
- Social-Emotional and Mental Components: Addressing the social impact (positive and negative) of new technologies, including aspects like information security, is crucial. These considerations are being integrated into statewide and national workforce strategic plans (e.g., National Governor's Association).
- Education System Integration: K-12 and post-secondary education are increasingly reinforcing these skills in youth.
- Vendor Engagement: There's an opportunity for major tech vendors (Amazon, Microsoft, Google) to further engage in workforce development and education to prepare individuals who are adaptable to future technologies.
Energy Consumption and Sustainability in AI
A notable point raised is the significant energy consumption of AI tools and data centers. Terracon is actively exploring ways to reduce the environmental cost of AI processes:
- Optimizing AI Workflows: For tasks like extracting content for project capsules or resumes, instead of repeatedly running AI on the entire data trove, information can be extracted once and staged in a data store. This allows for consumption at a much lower environmental cost.
- Partnership with Sustainability Teams: Terracon aims to partner with sustainability teams to integrate environmental considerations into AI process development.
Conclusion and Advice for AEC Leaders
David Harwood offers encouraging advice for AEC leaders:
- Don't feel behind: Firms, even the smallest, can leverage AI to advance core services.
- Focus on personal productivity first: Achieving early, meaningful success through personal efficiency improvements can be a great starting point while enterprise governance is being established.
- Incremental improvements matter: In a billable hours model, even small gains in personal efficiency can lead to significant organizational benefits.
He emphasizes that while the change management can be challenging, the benefits are substantial. David Harwood can be reached via email at david.harwood@terracon.com for further questions.
Chat with this Video
AI-PoweredHi! I can answer questions about this video "What Does it Take for AEC Firms to Successfully Adopt SCALABLE AI?". What would you like to know?