Yogashri Pradhan， SPE（高级石油工程师），现任OPX Ai首席增长官。她曾任雪佛龙公司首席生产工程师，在米德兰盆地和特拉华盆地拥有十余年非常规油气资产开发和生产工程经验。她同时也是IronLady Energy Advisors的创始人，该公司是一家专注于能源领域技术解决方案的咨询公司。Pradhan曾荣获德克萨斯大学奥斯汀分校石油与地球系统工程系杰出校友称号，并入选Hart Energy“40位40岁以下精英”奖项。她荣获2020年SPE西南北美区域油藏描述与动力学奖和区域服务奖、2018年SPE国际青年会员杰出服务奖，并于2018年被SPE墨西哥湾沿岸分会评为年度青年工程师。普拉丹拥有德克萨斯大学奥斯汀分校石油工程学士学位、德克萨斯农工大学石油工程硕士学位以及芝加哥大学布斯商学院工商管理硕士学位。她是德克萨斯州和新墨西哥州的注册专业工程师。
”
Jai Joon是OPX Ai的创始人兼首席执行官，该公司致力于通过人工智能解决方案，将复杂的运营环境与可执行的智能信息相结合。在创立这家软件开发公司之前，Joon曾在雪佛龙和康菲石油担任工程职务，领导过多个数字化转型项目。他是OPX Ai“集成运营中心即服务”（IOCaaS）模式以及面向北美油井监测和生产优化的AI驱动战略的主要架构师。
Case Study: Field Deployments of AI-Based IOCaaS Advancing Artificial Lift and Flow Assurance
Examples demonstrate how an Integrated Operations Center as a Service (IOCaaS) model, powered by artificial intelligence, reduced costs by 5% and increased production by 6% in Canada.
Oil wells in Canada.
Source: Getty Images.
Digital transformation in upstream operations has long promised greater efficiency and uptime, but traditional efforts often stall due to siloed systems and expensive infrastructure overhauls. Integrated Operations Center as a Service (IOCaaS) represents a new approach to enable the rapid adoption of an artificial intelligence (AI)-enabled operations model that represents an alternative to physical control rooms or large CAPEX investments.
IOCaaS provides a remote, cloud- or edge-hosted operations center that leverages existing field data streams and domain expertise. Early deployments of IOCaaS in North America are yielding results that include lower operating costs and higher production. This has been achieved by combining advanced analytics with real-time field data.
This case study shows how the IOCaaS, as developed by OPX Ai, is applied in oil and gas fields to optimize artificial lift and flow assurance. The discussion examines key technical aspects, case histories from Chevron and ConocoPhillips assets in Canada, implementation timelines, and quantifiable outcomes achieved. The focus is on how self-learning models and edge microservices are deployed, how they integrate with supervisory control and data acquisition (SCADA) and data historian systems, and what results they have delivered in the field.
IOCaaS Architecture and Approach
IOCaaS reimagines the traditional operations center as a service layer that sits atop an operator’s existing automation and IT infrastructure. Instead of building a centralized physical control room, operators subscribe to modular microservices (either cloud-connected or deployed at the edge) that continuously monitor and optimize assets.
These microservices interface with existing SCADA systems, historians, and enterprise resource planning (ERP) databases, i.e., there is no need to “rip-and-replace” legacy systems. Data from wells, compressors, pipelines, and facilities flow into the IOCaaS platform, where AI-driven analytics turn raw readings into actionable insights.
This architecture (Fig. 1) typically involves: field data acquisition from sensors and SCADA, an edge-computing layer for real-time analytics, cloud dashboards for engineers, and integration hooks into maintenance and business planning systems. By building on existing field infrastructure, IOCaaS enables the activation of integrated operations in a matter of weeks to a few months, rather than years.
Fig 1—Schematic of an AI-driven Integrated Operations Center as a Service (IOCaaS) architecture, showing how edge-deployed microservices interface with existing SCADA/historian data sources and cloud analytics to enable real‑time, exception-based management of wells and facilities.
Source: OPX Ai.
Critically, IOCaaS is designed for exception-based workflows. Instead of personnel watching screens 24 hours a day, 7 days a week to catch problems, the system monitors every well and piece of equipment autonomously. Engineers and operators are alerted only when anomalies or suboptimal conditions are detected, allowing teams to shift from reactive firefighting to proactive surveillance.
In Chevron’s Kaybob Duvernay Formation operation and ConocoPhillips’ Montney Formation asset, this translated to a leaner operating model. Field staff could take their eyes off routine wells and focus on higher-value tasks, confident that the AI will flag the exceptions. The end goal was a hybrid intelligence approach that relied on human expertise with guidance from AI analysis. This framework improved both the speed and quality of operational decisions.
Chevron’s Kaybob Duvernay IOCaaS Deployment
Chevron’s Kaybob Duvernay development in Alberta provided a testing ground for IOCaaS in a large, mature asset with legacy infrastructure.
Starting in 2020, Chevron partnered with OPX Ai to deploy IOCaaS across dozens of gas wells, associated compressors, and a central processing facility. The deployment was phased over about 12 months, aligning with the operator’s cautious approach to new technology in a live asset. In the first phase, data pipelines were established from field SCADA and historians into the cloud-based IOCaaS platform.
This included real-time wellhead pressures, compressor statuses, plunger lift cycles, and even maintenance work orders from Chevron’s ERP, all of which were integrated to give the AI a holistic view of operations. The software platform developer’s domain experts worked closely with the operator’s production engineers to configure the AI models to site-specific conditions, e.g., tailoring the hydrate model to the field’s gas composition and pipeline network. By early 2022, the IOCaaS was fully live, effectively functioning as a remote operations center for the asset.
In terms of process changes, the operator achieved a quick shift to exception-based management. Daily production calls and surveillance meetings were augmented with IOCaaS dashboards, highlighting only the wells and facilities that deviated from expected behavior.
For example, one morning the AI flagged an anomalous drop in plunger travel velocity on a single well, a subtle sign of liquid loading. This prompted an operator to intervene hours before that well would have accumulated enough fluid to trigger an automatic shutdown. This type of preemptive action became routine.
The AI artificial lift optimizer also systematically trimmed plunger gas injection on many wells, saving fuel and compressor runtime. Over the first year, the operator recorded a 5% reduction in lease operating expenses (LOE) in the Kaybob asset, attributable largely to lower fuel-gas usage, fewer callouts for hydrate-plug remediation, and fewer unplanned well interventions. By automating routine optimization and preventing problems, IOCaaS lowered the cost of each barrel of oil produced.
In terms of production performance, the field’s decline curve was shallower than in recent prior years. After normalizing for new-drill wells, BOE output was about 6% higher than originally projected, an uplift partly credited to the continuous lift optimizations keeping wells at their ideal operating point longer than was previously possible.
While external factors (i.e., commodity prices and facility debottlenecking) also influenced economics, the oil company’s internal analysis indicated measurable incremental value from the IOCaaS implementation, including the avoidance of an estimated 71,000 BOE of deferred production and several million dollars in cost savings over a 12-month period.
In addition to the quantitative results, the operator’s field and office teams developed a culture that embraced AI-generated recommendations. During the implementation phase, the operator kept humans in the loop by requiring engineers to manually review and approve each AI-recommended action.
As the models proved their accuracy over time, more decisions were automated. By the end of the deployment year, the operator allowed the IOCaaS to autonomously adjust certain gas lift valve set points and regenerate compressor restarts (within safety limits) without direct human approval.
This progressive handover illustrates a pathway to true autonomy, achieved through building confidence in AI.
Chevron’s experience also highlighted integration challenges, including the mapping of the myriad SCADA tags and cleaning years of historical data for training. Technical hurdles required effort but did not require operations to be paused. Because the IOCaaS overlays running systems, much of the setup and training occurred in parallel with normal production.
ConocoPhillips’ Montney Asset IOCaaS Deployment
In contrast to Chevron’s phased implementation, ConocoPhillips adopted the IOCaaS model in its Montney unconventional asset within a comparatively short timeframe. The Montney project, involving a multiwell pad development in British Columbia, progressed from kickoff to full deployment within 4 months.
Several factors enabled this rapid timeline. First, the operator had the advantage of “greenfield” digital infrastructure. The Montney facilities were newer and already had modern, standardized SCADA and data historians, making integration simpler. Second, the software developer leveraged the configurations and lessons learned from Kaybob and other projects, essentially deploying a templated solution customized to the operator’s needs. As a result, the Montney operations team had a functioning IOCaaS portal overseeing its wells and gathering network within about 120 days.
The focus areas in Montney were slightly different. While artificial lift optimization (i.e., gas lift) was used on initial Montney wells, the more immediate concern was flow assurance and facility uptime in a harsh winter climate. The IOCaaS was set to monitor for hydrate conditions, compressor performance, and liquid loading in pipelines.
In January of the project’s first winter, when temperatures plummeted, the system’s hydrate-risk model delivered value by alerting the team to inject methanol at two critical chokepoints. This prevented what would likely have been freeze-ups. Field operators reported that normally they do not necessarily pre-emptively dose at those points, since surface signs of hydrates weren’t yet obvious.
In contrast, the AI pattern-recognition model caught the subtle signs of minor pressure fluctuations and cooling trends that may not have been apparent through human oversight. Additionally, the rotating equipment (i.e., gas compressors) was under AI surveillance for early failure warnings.
On at least one occasion, the AI detected an anomaly in a compressor’s vibration signature and discharge pressure trend that led to a controlled shutdown for inspection. This averted a run-to-failure scenario that could have caused days of downtime.
After 4 months of operation, ConocoPhillips saw measurable benefits. Even with only a partial year of data, the Montney asset team calculated a 3 to 4% production increase above forecast on the AI-optimized wells.
The operator also anticipated it would achieve approximately a 6% uplift as more wells were brought online under IOCaaS optimization. Downtime was significantly reduced, as no hydrate-related outages occurred during the evaluation period. This contributed to an overall reduction in LOE of approximately 5%, supported by fewer emergency callouts and more efficient chemical usage.
The early success led ConocoPhillips to expand IOCaaS coverage to more pads and to evaluate its potential application in another Canadian asset.
The Montney deployment demonstrated the scalability of the approach. By delivering the solution through a service model, the software developer replicated the functionality of a conventional operations center, which typically requires about 1 year to establish, and deployed it in a new asset in roughly one-third of that time. This accelerated implementation provides a potential framework for other operators seeking to achieve similar outcomes without extended project timelines.
Conclusions
The field deployments at Chevron and ConocoPhillips demonstrate that the IOCaaS model has advanced beyond the conceptual stage to become a practical solution delivering measurable operational improvements. Across the initial implementation cycles, the following key outcomes were observed:
Approximately 5% reduction in LOE was achieved through efficiency gains, primarily from reduced fuel-gas consumption in artificial lift systems, optimized chemical injection (i.e., applying inhibitors only when required), and fewer unplanned interventions.
Average BOE production increased by roughly 6% through continuous well optimization, maintaining wells closer to their ideal operating points than was previously feasible. The cumulative effect across large well inventories represents significant incremental production without new drilling activity.
Proactive identification of potential failures and flow interruptions prevented an estimated 71,000 BOE of deferred production. Early detection of issues such as hydrate formation and electrical submersible pump failures helped protect millions of dollars in potential revenue while improving asset reliability and operational safety.
Many of the value gains realized were achieved without major capital projects or new hardware installations. The IOCaaS framework leverages existing field sensors and equipment, with performance improvements derived primarily from enhanced data integration and analytics. These outcomes indicate a strong return on investment, an important consideration in the current cost-sensitive oil and gas environment.
Additionally, instead of adding more dashboards, and by focusing on filtering information by exceptions, IOCaaS improved surveillance efficiency by up to 30% in separate pilots. Taken together, the results demonstrate that the AI-based technology enabled engineers to manage a greater number of wells per person, supporting asset growth despite ongoing workforce shortages in the oil and gas industry.
Yogashri Pradhan, SPE, is chief growth officer at OPX Ai. She previously worked as a lead production engineer at Chevron and has more than a decade of experience in unconventional asset development and production engineering across the Midland and Delaware basins. She is also the founder of IronLady Energy Advisors, a consulting firm focused on technical solutions across the energy spectrum. Pradhan has been recognized as a distinguished alumna of The University of Texas at Austin’s Department of Petroleum and Geosystems Engineering and was named to Hart Energy’s 40 Under 40 award program. She received the 2020 SPE Southwestern North America Regional Reservoir Description and Dynamics and Regional Service Awards, the 2018 SPE International Young Member Outstanding Service Award, and was named Young Engineer of the Year by the SPE Gulf Coast Section in 2018. Pradhan holds a BSc in petroleum engineering from The University of Texas at Austin, an MS in petroleum engineering from Texas A&M University, and an MBA from the University of Chicago Booth School of Business. She is a licensed professional engineer in Texas and New Mexico.
Jai Joon is the founder and CEO of OPX Ai, which aims to connect operational complexity with actionable intelligence through AI-based solutions. Before founding the software developer, Joon held engineering roles at Chevron and ConocoPhillips, where he led several digital transformation initiatives. He is the principal architect of OPX Ai’s Integrated Operations Center as a Service (IOCaaS) model and AI-driven strategies for well surveillance and production optimization across North America.
