Bentley is adding MCP servers across its engineering software so AI agents can automate structural analysis, modeling, monitoring, and repetitive tasks while engineers retain control of technical decisions.
Structural engineering faces mounting pressure from talent shortages, growing infrastructure demand, tighter sustainability requirements, and increasingly complex projects. Bentley Systems argues that AI can help engineers respond by automating routine work while preserving the judgment and accountability required in safety-critical design.
Bentley’s strategy centers on Model Context Protocol, or MCP, an open standard that connects AI agents directly to engineering applications. Engineers can describe tasks in natural language rather than navigate interfaces or write scripts. The AI interprets the request, while Bentley software executes deterministic functions, providing reproducible engineering results and reducing the risk associated with AI hallucinations.
STAAD.Pro demonstrates this approach in structural analysis. Its MCP server lets agents apply loads, revise models, assign member properties, run analyses, and compare results through conversational instructions. Engineers remain responsible for interpreting results and deciding what to do next. STAAD Foundation Advanced extends the concept to foundation design, while SACS applies it to offshore structural modeling and analysis.
Bentley is also extending MCP beyond structural analysis. MicroStation agents can interact with elements, levels, references, and views in DGN files, while iTwin IoT enables conversational access to live and historical sensor data. This could make structural health monitoring more accessible by helping engineers investigate alerts, analyze trends, and generate reports without specialized data science skills.
Bentley sees further opportunities in design exploration. Machine-learning models trained on previous simulations could rapidly estimate structural behavior across thousands of configurations before engineers perform high-fidelity analyses on promising candidates. Generative design could similarly use historical project information to propose alternatives and reuse proven approaches.
The company advocates gradual adoption rather than wholesale automation. Firms can begin with bounded, verifiable tasks and expand as confidence grows. In this model, AI handles more repetitive execution while structural engineers concentrate on judgment, optimization, creativity, safety, and professional responsibility.
