Corva and Oliasoft Build a Continuous Learning Loop for Smarter Well Construction Introduction Corva AI and Oliasoft have announced a strategic collaboration designed to connect well engineering, real-time drilling operations and post-well analysis through a bidirectional digital workflow. The integration links Oliasoft WellDesign with Corva’s Real-Time AI platform, creating a continuous information loop in which engineering plans guide field execution and drilling results improve future well designs. The partnership addresses one of the persistent weaknesses in modern well construction: the separation between planning software and operational systems. Although operators generate large volumes of engineering and drilling data, those datasets often remain fragmented across departments, applications and reporting processes. The result is slower decision-making, repeated manual work and limited transfer of lessons from one well to the next. By connecting design assumptions with liv...
Introduction
.jpg)
Corva AI and Oliasoft have announced a strategic collaboration designed to connect well engineering, real-time drilling operations and post-well analysis through a bidirectional digital workflow. The integration links Oliasoft WellDesign with Corva’s Real-Time AI platform, creating a continuous information loop in which engineering plans guide field execution and drilling results improve future well designs.
The partnership addresses one of the persistent weaknesses in modern well construction: the separation between planning software and operational systems. Although operators generate large volumes of engineering and drilling data, those datasets often remain fragmented across departments, applications and reporting processes. The result is slower decision-making, repeated manual work and limited transfer of lessons from one well to the next.
By connecting design assumptions with live rig data, Corva and Oliasoft aim to create what they describe as a digital thread across the well lifecycle. The model begins with engineering intent, moves through real-time execution, captures observed performance and feeds the results back into future planning.
The initiative is primarily aimed at oil and gas well construction, but its underlying architecture could also have implications for geothermal, carbon capture and storage, and other industries that rely on complex subsurface wells.
Bridging Planning and Operations
Well construction typically involves several stages, including trajectory design, casing planning, hydraulics analysis, drilling mechanics, anti-collision checks, execution and post-well review. Each stage produces important information, but the information is not always available to the people who need it at the right time.
A drilling engineer may create a well plan using assumptions about formation pressure, mud properties, torque, drag, hole cleaning and wellbore stability. Once drilling begins, however, the rig generates continuously changing data. Actual hookloads, rates of penetration, pump pressures, mud properties, vibration, torque and wellbore position may differ from the original model.
Without a direct connection between the design system and the operational environment, teams may have to export data manually, reformat files or reconstruct engineering assumptions from reports. That process can make it difficult to determine whether a change in drilling behavior represents a normal variation, a design-limit issue or a developing operational risk.
The Corva–Oliasoft integration is intended to reduce this gap. Planned trajectories, casing designs, pressure profiles and related engineering parameters can move from Oliasoft WellDesign into Corva’s operational platform. Drilling teams can then view live activity in the context of the original plan rather than relying only on isolated rig measurements.
This distinction is important. A live data point has limited value without context. For example, a torque increase may be significant if it approaches a modeled limit, but less concerning if it remains within the expected range for a particular hole section. Similarly, a deviation from the planned trajectory may require immediate attention if it affects anti-collision separation or casing placement.
From Static Plans to Live Engineering Context
Traditional well planning often produces a document or set of engineering outputs that is treated as the operational reference for the well. In practice, the plan evolves as conditions change. Formation behavior, equipment performance, drilling parameters and operational constraints can all force modifications.
The challenge is ensuring that changes made during drilling are captured in a structured way and remain visible to the engineering team. If field adjustments are recorded only in daily reports, emails or informal conversations, valuable information can be lost. Future well plans may then repeat the same assumptions without accounting for what actually happened.
The planned integration is intended to provide continuous visibility between the well model and field execution. Corva’s real-time environment can receive the design intent from Oliasoft, while operational results can be returned to the well engineering platform.
This creates a plan-versus-actual workflow. Teams can compare:
- Planned and as-drilled trajectories.
- Modeled and observed pressure behavior.
- Expected and actual torque and drag.
- Designed and achieved casing depths.
- Planned drilling parameters and field performance.
- Forecast operational duration and actual time spent.
- Anticipated and recorded nonproductive-time events.
The goal is not merely to transfer files between two applications. It is to preserve the relationship between an engineering decision and the field evidence that confirms, challenges or refines that decision.
A Feedback Loop Across Wells
The central concept behind the collaboration is a continuous learning loop:
Engineering intent → real-time execution → observed performance → model calibration → improved next well
In a conventional workflow, learning may occur during a post-well review weeks after the rig has moved on. By then, operational details may be incomplete, data may be stored in different formats and the people involved in the original decisions may no longer be available.
A connected workflow can make learning more structured. As-drilled trajectories, drilling data and nonproductive-time events can be returned to Oliasoft and used to calibrate future engineering models. This allows planning teams to compare assumptions with actual field performance and adjust subsequent designs.
For operators drilling multiple wells in the same field or basin, the potential value is substantial. Repeated well programs generate patterns that may not be visible in a single operation. A particular casing design may perform better in one formation interval. A specific drilling parameter range may reduce vibration. A recurring pressure response may indicate that the original hydraulic model requires revision.
When these observations are captured consistently, they can become program-level knowledge rather than isolated experience held by individual engineers or crews.
Plan-Aware Operations
One of the main workflows supported by the integration is plan-aware operations. This means giving drilling personnel access to the engineering information behind the well plan while the well is being drilled.
The approach can help teams understand whether actual operations remain within defined limits. Instead of viewing trajectory, pressure or drilling parameters as separate datasets, the team can interpret them against the assumptions that shaped the design.
Plan-aware operations may be particularly valuable in complex wells, where small deviations can create significant downstream consequences. Extended-reach wells, high-angle wells, deepwater wells and unconventional laterals often involve narrow operating windows. A seemingly minor change in wellbore position or mechanical behavior can affect casing installation, completion access or collision risk.
The integration could also improve communication between office-based engineers and rig personnel. Both groups can work from a more consistent reference point, reducing the risk that one team is assessing a current operation against an outdated version of the plan.
Engineering-Contextualized Monitoring
Real-time monitoring systems can display thousands of data points. However, more data does not automatically produce better decisions. Engineers and drillers need to identify which changes matter and why.
Engineering-contextualized monitoring attempts to address this issue by combining live measurements with modeled expectations and operating limits. Actual hookloads, mud properties, drilling parameters and wellbore position can be assessed in relation to the design.
For example, an increase in torque may be interpreted differently depending on the modeled torque-and-drag profile, current hole depth, trajectory and planned bottomhole assembly. A change in pump pressure may require a different response depending on hydraulics calculations and the expected pressure regime.
This context can make real-time systems more useful for decision support. Instead of simply alerting users that a parameter has changed, the system can help indicate whether the change is relevant to the well design and whether it warrants investigation.
That does not eliminate the need for experienced personnel. Rather, it can help engineers focus their attention on deviations with the greatest potential impact.
Live Engineering Analysis
The integration is also designed to keep operational inputs current for engineering analyses such as torque and drag, hydraulics, hole cleaning and anti-collision.
These analyses are often performed before drilling begins and then updated as the well progresses. If the operational data needed for updates is difficult to access, models can become detached from actual conditions.
A more direct data connection could allow engineers to assess how the well is performing against the model during critical phases. In a directional drilling operation, for example, updated trajectory information can support anti-collision analysis and help identify whether the well remains within its planned corridor.
In hydraulics and hole-cleaning analysis, current drilling parameters and mud properties can improve the relevance of calculations. In torque-and-drag analysis, the as-drilled trajectory and measured mechanical response can provide evidence for evaluating friction assumptions and operational limits.
The value of these capabilities will depend on data quality, model validation and the speed at which information moves between the field and office. Nevertheless, maintaining current operational inputs is an important step toward making engineering analysis more responsive.
Capturing Nonproductive Time
Nonproductive time is one of the most important sources of operational learning in drilling. It can result from stuck pipe, lost circulation, equipment failure, well-control events, excessive connection time, hole-cleaning problems, unplanned trips or other issues.
However, NPT data is often difficult to compare across wells. Different teams may classify events differently, while reports may lack the engineering context needed to identify the underlying cause.
By returning NPT events and operational data to the well design environment, the proposed workflow could help operators connect time losses to specific planning assumptions or execution conditions. Over multiple wells, this may reveal recurring patterns.
For instance, an operator could determine that a particular well section consistently experiences elevated drag, that a casing depth is difficult to reach under certain pressure conditions or that a specific drilling parameter combination is associated with vibration. Such findings could inform the design of future wells before the rig begins work.
Reducing NPT is not only an operational objective. It can also improve project economics by lowering rig time, reducing service costs and increasing confidence in well delivery schedules.
Relevance to Geothermal Drilling
Although the announcement is centered on oil and gas well construction, the same digital thread could be relevant to geothermal development. Geothermal wells often face demanding temperature, pressure, rock-strength and drilling-environment conditions. Projects may also have limited well-count economics, making each well particularly important.
In conventional geothermal development, the cost and risk of drilling can represent a major share of total project expenditure. A connected well-planning and execution workflow could help developers preserve learning across production, injection and appraisal wells.
The technology may be useful for tracking:
- High-temperature drilling performance.
- Lost circulation events.
- Wellbore stability and borehole enlargement.
- Casing and cementing performance.
- Directional trajectory accuracy.
- Drill-string torque and drag.
- Temperature and pressure conditions.
- Time spent on trips, repairs and remedial work.
The value of cross-well learning may be especially important for emerging geothermal projects, including enhanced geothermal systems. EGS wells can involve complex trajectories, hard crystalline rock and demanding stimulation objectives. Better integration between design assumptions and field outcomes could support more reliable planning as developers move from demonstration wells toward repeatable commercial programs.
The press release does not announce a dedicated geothermal deployment or customer result. However, Oliasoft states that its software serves oil and gas, geothermal and carbon-capture companies, making the broader architecture potentially relevant to multiple subsurface industries.
Implications for CCS and Multi-Energy Wells
Carbon capture and storage projects also depend on well engineering, subsurface characterization and long-term integrity. Injection wells may require careful management of pressure, casing, cement, temperature and formation response.
A connected workflow could help CCS developers compare planned injection-well designs with actual drilling conditions and preserve field evidence for future wells. This may become increasingly important as projects expand from individual demonstration wells to multi-well storage hubs.
The collaboration also reflects a broader trend toward multi-energy well engineering. The same software infrastructure can support wells used for hydrocarbon production, geothermal heat, carbon storage and potentially other subsurface applications.
This convergence creates an opportunity for software platforms to provide common engineering and operational workflows across energy sectors. It may also allow companies to transfer lessons between industries, although differences in well objectives, regulatory requirements and operating conditions must be carefully managed.
AI Needs Engineering Context
The announcement highlights an important principle in industrial artificial intelligence: AI systems require reliable context, not just large volumes of data.
A model trained on drilling data may identify patterns in torque, pressure, rate of penetration or vibration. But interpreting those patterns requires knowledge of the well design, formation, equipment, operating limits and engineering objectives.
Connecting Corva’s real-time operational environment with Oliasoft’s engineering models could provide that context. Live drilling intelligence would be linked to the assumptions and constraints that define acceptable performance.
This may support more meaningful alerts, better deviation analysis and stronger recommendations. It could also help prevent a common problem in industrial AI: generating technically plausible outputs that are disconnected from the actual decision being made.
The companies’ approach appears to position AI as part of a broader engineering workflow rather than as a replacement for engineering judgment. Physics-based models provide the technical foundation, while real-time data and AI help teams interpret changing conditions.
Implementation Challenges
The potential benefits of the integration will depend on execution. Connecting two platforms is not enough if data definitions, units, event classifications or well identifiers are inconsistent.
Operators will need robust data governance covering:
- Common naming conventions.
- Version control for well plans.
- Reliable time synchronization.
- Clear ownership of engineering inputs.
- Consistent classification of NPT.
- Secure data exchange between systems.
- Validation of automated calculations.
- Human review of model changes.
There is also an organizational challenge. A continuous learning loop requires teams to record decisions and outcomes consistently. If operational events are poorly documented, the system may return incomplete or misleading information to the planning environment.
Cybersecurity and access control will also be important. A platform connected to live drilling operations must protect sensitive operational data and ensure that changes to engineering models are authorized and traceable.
The initial customer pilots will therefore be important. They should demonstrate not only that data can move between systems, but also that the integration improves decision speed, reduces manual reconciliation and produces measurable operational value.
What the Partnership Signals
The Corva–Oliasoft collaboration reflects a shift in the digitalization of well construction. Earlier software initiatives often focused on digitizing individual tasks or replacing paper-based workflows. The next stage is connecting those tasks across the full lifecycle of a well.
This means treating the well plan as a living engineering model rather than a static document. The model informs operations, operations update the model and the resulting knowledge improves future designs.
For operators, the strategic benefit could extend beyond individual wells. A company that captures engineering and operational learning systematically may improve consistency across rigs, assets and teams. That could support lower drilling costs, reduced risk and more predictable project delivery.
For geothermal and CCS developers, the same concept offers a path toward industrializing complex well programs. As these sectors seek to scale, the ability to learn quickly from each well may become as important as the ability to design the first well correctly.
The collaboration is still moving toward customer pilots, so its ultimate impact will depend on deployment results. Even so, the announcement points toward a future in which well engineering, real-time drilling intelligence and post-well learning operate as parts of one connected system rather than separate stages.
Source: Official Company Press Release
Comments
Post a Comment