Blab AI
DataRobot Partners with Chevron to Deploy Edge-Based Agentic AI for Autonomous Facility Inspections
Imagine a pipeline inspection that never stops checking itself. That’s the vision behind DataRobot’s new partnership with Chevron U.S.A., a subsidiary of Chevron Corporation, which will bring agentic artificial intelligence (AI) to the edge of the oil‑and‑gas giant’s facilities.
Under Chevron’s Facilities and Operations of the Future initiative, the collaboration aims to streamline how robotic missions are planned, assessed, and executed. Traditionally, each aerial or terrestrial robot that flies over a refinery or walks along a pipeline must first pass a lengthy operator‑verification process. The new system replaces that manual gatekeeping with advanced optimizing and reasoning agents that generate mission plans on the fly and continuously monitor operating conditions.
DataRobot’s Agent Workforce Platform stitches together specialized agents—such as sensor‑analysis models and geospatial reasoners—into a coordinated workflow that plugs directly into Chevron’s existing operational systems. Because the platform runs at the edge, it does not require a wholesale infrastructure overhaul; teams can evaluate AI capabilities right where the work happens.
A core component is the Safe Start agentic assessment, powered by NVIDIA NIM microservices. This assessment evaluates conditions before and during robotic missions by fusing hard‑wired gas sensors, supplemental vision systems, and AI models to keep a real‑time safety net around each asset.
"Agentic AI lets us shift from manually approving each mission to continuously assessing conditions in real time," said Debanjan Saha, CEO of DataRobot. "With the Agent Workforce Platform, Chevron can deploy AI decision‑making at the edge while still meeting stringent security, reliability, and governance requirements."
"In simple terms, this improves safety by reducing reliance on manual, point‑in‑time checks and validating conditions continuously while work is underway," added Cari Armpriester, Facilities and Operations of the Future program manager at Chevron. "It helps reduce exposure, apply safety standards consistently, and deploy robotics more efficiently across our operations."
By moving the focus from hardware certification to environmental condition assurance, the partnership promises fewer human interventions and more consistent deployment of robotic inspections—without sacrificing oversight.
The initiative leverages NVIDIA’s AI software and compute stack within Chevron’s digital and operational ecosystems. The DataRobot platform orchestrates the agents so they can operate in the real‑time, safety‑critical environments typical of oil and gas facilities.
Edge AI is already reshaping industrial operations: it cuts latency, keeps sensitive data on site, and enables continuous monitoring without relying on cloud connectivity. Agentic AI—systems that can pursue goals, reason, and act autonomously—offers a more adaptive approach to mission planning than static, pre‑approved routes.
Chevron’s existing robotic fleet, which includes drones and ground robots that inspect pipelines, storage tanks, and other critical infrastructure, will now be able to adjust paths and actions on the fly based on live sensor data. This capability could slash inspection cycles and reduce the need for human intervention.
The platform is built to integrate with a variety of AI models and data sources. By combining sensor analysis, geospatial reasoning, and safety assessment agents, it produces a comprehensive plan that satisfies operational standards while responding to changing conditions.
Governance and security remain top priorities. The platform’s design incorporates safeguards that align with Chevron’s security and reliability requirements, ensuring autonomous decisions are traceable and compliant with industry regulations.
Although still in the pilot phase, the companies have already begun testing the system in controlled environments. Future plans include scaling the solution across more Chevron sites and expanding the range of robotic missions that can benefit from continuous, AI‑driven assessment.
The partnership underscores how AI can be integrated into existing industrial workflows without replacing legacy infrastructure and illustrates the growing role of agentic AI in enhancing safety, efficiency, and reliability in high‑stakes environments.
For more information about the collaboration, visit www.datarobot.com.
Under Chevron’s Facilities and Operations of the Future initiative, the collaboration aims to streamline how robotic missions are planned, assessed, and executed. Traditionally, each aerial or terrestrial robot that flies over a refinery or walks along a pipeline must first pass a lengthy operator‑verification process. The new system replaces that manual gatekeeping with advanced optimizing and reasoning agents that generate mission plans on the fly and continuously monitor operating conditions.
DataRobot’s Agent Workforce Platform stitches together specialized agents—such as sensor‑analysis models and geospatial reasoners—into a coordinated workflow that plugs directly into Chevron’s existing operational systems. Because the platform runs at the edge, it does not require a wholesale infrastructure overhaul; teams can evaluate AI capabilities right where the work happens.
A core component is the Safe Start agentic assessment, powered by NVIDIA NIM microservices. This assessment evaluates conditions before and during robotic missions by fusing hard‑wired gas sensors, supplemental vision systems, and AI models to keep a real‑time safety net around each asset.
"Agentic AI lets us shift from manually approving each mission to continuously assessing conditions in real time," said Debanjan Saha, CEO of DataRobot. "With the Agent Workforce Platform, Chevron can deploy AI decision‑making at the edge while still meeting stringent security, reliability, and governance requirements."
"In simple terms, this improves safety by reducing reliance on manual, point‑in‑time checks and validating conditions continuously while work is underway," added Cari Armpriester, Facilities and Operations of the Future program manager at Chevron. "It helps reduce exposure, apply safety standards consistently, and deploy robotics more efficiently across our operations."
By moving the focus from hardware certification to environmental condition assurance, the partnership promises fewer human interventions and more consistent deployment of robotic inspections—without sacrificing oversight.
The initiative leverages NVIDIA’s AI software and compute stack within Chevron’s digital and operational ecosystems. The DataRobot platform orchestrates the agents so they can operate in the real‑time, safety‑critical environments typical of oil and gas facilities.
Edge AI is already reshaping industrial operations: it cuts latency, keeps sensitive data on site, and enables continuous monitoring without relying on cloud connectivity. Agentic AI—systems that can pursue goals, reason, and act autonomously—offers a more adaptive approach to mission planning than static, pre‑approved routes.
Chevron’s existing robotic fleet, which includes drones and ground robots that inspect pipelines, storage tanks, and other critical infrastructure, will now be able to adjust paths and actions on the fly based on live sensor data. This capability could slash inspection cycles and reduce the need for human intervention.
The platform is built to integrate with a variety of AI models and data sources. By combining sensor analysis, geospatial reasoning, and safety assessment agents, it produces a comprehensive plan that satisfies operational standards while responding to changing conditions.
Governance and security remain top priorities. The platform’s design incorporates safeguards that align with Chevron’s security and reliability requirements, ensuring autonomous decisions are traceable and compliant with industry regulations.
Although still in the pilot phase, the companies have already begun testing the system in controlled environments. Future plans include scaling the solution across more Chevron sites and expanding the range of robotic missions that can benefit from continuous, AI‑driven assessment.
The partnership underscores how AI can be integrated into existing industrial workflows without replacing legacy infrastructure and illustrates the growing role of agentic AI in enhancing safety, efficiency, and reliability in high‑stakes environments.
For more information about the collaboration, visit www.datarobot.com.
