ASCE's AI RACE Roadmap: What Civil Engineers Should Know in 2026

ASCE is moving artificial intelligence from scattered experimentation toward a more coordinated professional strategy for civil engineering.

The key development is AI RACE—a roadmap being developed to organize where AI can add value, how it should be scaled, and what oversight should remain in place as adoption accelerates.

ASCE describes AI RACE as the AI Roadmap for Application in Civil Engineering. The proposed roadmap is intended to create a consistent vision for AI across civil engineering rather than leaving each discipline, firm, or research group to develop its own disconnected approach.

The announcement identifies several themes the roadmap is expected to address:

• AI applications and outcomes across civil engineering practice
• ethics and responsible use
• oversight and accountability
• strategies for preventing disciplinary and organizational silos
• methods for scaling AI technologies beyond isolated pilots
• short-, medium-, and long-term goals for implementation

ASCE also states that the roadmap is intended to clarify the leadership role the Society can play as AI affects civil engineering practice.

AI adoption in civil engineering carries a different risk profile from many general productivity uses because engineering decisions can affect public safety, infrastructure reliability, project cost, environmental performance, and long-term asset behavior.

The central question is therefore not simply whether AI can produce a faster answer. Engineers must also determine whether the result is traceable, technically valid, applicable to the project, reproducible, and reviewed by a qualified human.

A useful professional framing is:

AI may accelerate engineering work, but it does not transfer engineering responsibility away from the engineer.

That principle is explicit in ASCE Policy Statement 573 — Artificial Intelligence and Engineering Responsibility. The policy states that civil engineers retain responsibility for planning, design, construction, operations, maintenance, and protection of public health, safety, and welfare. AI cannot replace the professional judgment of a licensed Professional Engineer.

ASCE's AI RACE announcement is notable because it documents current applications rather than discussing AI only as a future possibility.

Structural engineering

ASCE cites current use of AI in:

• design automation
• structural assessment
• surrogate models for computationally complex simulations
• decision-making support
• risk assessment

A surrogate model can be useful when a high-fidelity analysis is expensive to run repeatedly, but it remains dependent on the quality and coverage of the data used to train or calibrate it. It should not be treated as a universal substitute for first-principles analysis.

Transportation engineering

The ASCE announcement describes machine-learning applications that:

• collect and improve roadway data
• identify moving objects
• predict movement direction
• support crash-risk reduction

It also highlights computer-vision tasks using transportation cameras and vehicle-perception systems for classification, detection, and tracking.

Construction and infrastructure monitoring

ASCE identifies camera-based applications for:

• construction monitoring
• road-surface monitoring
• extracting decision-support information from field imagery

The same broader technology stack can be combined with sensors, inspection records, digital twins, and asset databases, but the engineering value depends on verified data pipelines and clear human review gates.

Enterprise knowledge and workflow support

ASCE also notes that companies are integrating chatbot systems such as ChatGPT and Gemini at the enterprise level.

That matters because enterprise AI can extend beyond drafting into document review, research, internal knowledge retrieval, project controls, reporting, and workflow automation.

The more autonomous the AI workflow becomes, the more explicit its permissions, stop conditions, audit trail, and human approval gates should be.

A major risk in rapidly adopting technology is fragmentation. One team may build a transportation model, another may deploy a document assistant, another may create an inspection classifier, and another may automate design checks—with different assumptions, data controls, validation methods, and accountability rules.

ASCE's stated emphasis on preventing silos is therefore technically important. A useful roadmap should help organizations establish common expectations for at least five areas:

These are not merely IT questions. In safety-critical contexts, they are part of engineering quality management.

Because the final roadmap has not yet been released, the most useful approach is to define the questions engineers should ask when it is published.

ASCE's AI activity is expanding beyond a single roadmap project. In August 2026, the Society reported that ASCE Publications is preparing a new AI in Civil Engineering journal, with submissions expected to begin in early 2027.

ASCE's current AI topic hub also features applications involving infrastructure inspection, bridge assessment, land analysis, transportation, construction, and other civil engineering domains.

Taken together, these developments show that ASCE is treating AI as a sustained professional and research area rather than a temporary productivity trend.

This creates a useful signal for civil engineering education as well: future engineers will likely need competency not only in using AI tools, but in checking them.

AI literacy should complement—not replace—engineering fundamentals.

A student who can prompt an AI system but cannot check equilibrium, dimensions, hydraulic continuity, soil behavior, structural load paths, transportation logic, statistics, or code applicability is not better prepared for engineering practice.

Stronger AI capability can actually increase the need for technical fundamentals because AI can produce more work, more quickly, and with more persuasive presentation.

A robust curriculum should therefore teach students to:

• separate facts from assumptions
• verify sources against original documents
• reproduce consequential calculations independently
• identify invalid units, geometry, loads, boundary conditions, or governing cases
• communicate uncertainty and limitations
• preserve an audit trail for AI-assisted work
• understand that professional responsibility remains human

For Philippine practice, the useful question is how to apply the roadmap's ideas without confusing international professional guidance with local regulatory authority. AI RACE can inform engineering workflows, governance, and education, but local legal and professional requirements remain controlling.

The practical relevance is nevertheless high. The same AI capabilities discussed by ASCE can support Philippine work in areas such as:

• flood and rainfall analysis
• traffic and mobility monitoring
• bridge and structural-health monitoring
• slope and landslide screening
• construction progress and quality documentation
• pavement and road-condition assessment
• infrastructure asset management
• surveying and remote sensing
• disaster-response decision support

For local practice, the governing rule should remain straightforward: AI can assist the engineering process, but the responsible professional must still verify the basis, applicability, and result before consequential use.

The next major event is the publication of the complete AI RACE roadmap, which ASCE says is planned for fall 2026.

When that document is released, civil engineers should look for concrete guidance on:

• professional responsibility
• validation requirements
• data and model governance
• education and workforce development
• integration with engineering software and digital twins
• autonomous and agentic systems
• implementation sequencing
• measurable near-term actions

Until then, the September 14 announcement should be read as a verified preview of ASCE's direction—not as the final policy architecture.