top of page

WillCo Drilling Intelligence - Edition 25: Measure the Use, Not the Install

Writer: William B. Contreras
William B. Contreras
7 days ago
16 min read


🛢️ Top Story

Coverage, Not Capability: In Oman, the Wells Where Drilling Automation Ran Most Drilled Fastest

Published November 2 in the ADIPEC 2026 proceedings (SPE-235814-MS).


Three rigs, three Oman fields, one repeated design: the wells where optimization ran most of the section posted higher drilling rates, Halliburton authors report.



At campaign scale, one rig delivered 28 wells in 440 days against an authorized plan of 540 days, about 100 days ahead. Performance did not climb over the fifteen months and no small group of wells carried the average; at the well level, the authors single out two wells delivered ahead of plan, one of them with zero NPT.


Automated steering was engaged in about 90% of the steered sections in every field and on every rig, so the authors treat it as a constant. What varied from well to well was coverage: how much of each section the drilling performance optimization (DPO) workflow covered, the one variable under study.



12-1/4-in. section, Field A: three benchmark wells drilled on the same rig at 0%, 31% and 81% DPO coverage. The authors report a 35% higher drilling rate (ROP) from the lowest to the highest coverage well, with no worsening of downhole dynamics at high coverage.


6-1/8-in. section, Field C: three wells at 1 to 8% coverage clustered together, and three wells at 69 to 81% coverage averaged a 24% higher drilling rate than the low-coverage group. The authors note that the benefit does not build up in step with coverage.


Upper Shuaiba development drilling repeats one well design across many wells, yet two wells built to the same plan can take a very different number of days to drill; because the wells are delivered in campaigns, the authors note, that difference is paid for on every well. The workflow, run on three rigs across three of the five Upper Shuaiba fields in Oman, combines optimized BHAs, an RSS in the 12-1/4-in. and 6-1/8-in. sections, automated steering, DPO and real-time geosteering to hold thin carbonate targets.


The authors present the result as an association across commercial wells, not a controlled trial, with single wells as the high-coverage cases in the 12-1/4-in. section. They tie it to the approximately 70% coverage threshold reported for closed-loop optimization in published field studies, and the 24% sits close to a gain measured between automated and conventional sections on the Norwegian Continental Shelf.


Two-panel bar chart titled Coverage, not capability. Field A, 12-1/4-in. section, three benchmark wells on the same rig: 0% DPO coverage 112.2 ft/hr, 31% coverage 127.6 ft/hr, 81% coverage 151.6 ft/hr, plus 35% from lowest to highest coverage. Field C, 6-1/8-in. lateral: one well with the workflow switched off 61.4 ft/hr, three wells at 1 to 8% coverage 89.9 ft/hr, three wells at 69 to 81% coverage 111.9 ft/hr, plus 24% for the high-coverage group over the low-coverage group. Source: SPE-235814-MS
Figure 1. Coverage, not capability: higher drilling performance optimization (DPO) coverage is associated with higher ROP in the Oman case set. Commercial wells, not a controlled trial; the high-coverage 12-1/4-in. cases are single wells. Source: SPE-235814-MS, m/hr converted to ft/hr.

Follow-up to Edition 24: a mile a day. ADNOC Onshore and Corva authors report the 6-in. reservoir section of Well B delivered 5,280 ft in 24 hours, a first for the operator (SPE-235580-MS).



Across two wells the system saved more than 59 on-bottom hours, 20% against the top-quartile (P75) plan, valued by the authors at more than US$208,000. Seven of eight sections beat the P75 benchmark. It ran in advisory mode, recommending parameter setpoints from WITSML data within limits agreed with the drilling team, with separate models for RSS and PDM runs. Read Edition 24.


🎯 Your move: ask for coverage, not availability, by section, rig and crew, against that approximately 70% mark (WillCo analysis).



⚖️ Counterpoint: The Verification Burden

When the Automation Is an Agent, Someone Still Has to Check Its Work

Both papers published November 2 in the ADIPEC 2026 proceedings.


At SLB, well classification fell from 8 hours to 20 minutes per well; at ADNOC Offshore, every figure an agent publishes still passes a human-review gate.


SLB authors report that their own AI platform (SPE-236279-MS) covers both well planning and execution, applies a suite of AI and ML engines to DDRs, including a transformer-based model that classifies drilling activities from operational text, and is used by more than 250 engineers, with 38% annual growth. Their reported results:

  • 💰 SLB estimates about $34.7 million a year in value from automated technology identification.

  • ⏱️ Well classification time: 8 hours to 20 minutes per well.

  • 📑 MENA tender: completed in at least half the previously required time, while evaluating all available offset wells.

  • 📄 Data preparation effort reduced, and end-of-well reports from four days to two.


ADNOC Offshore and Corva authors (SPE-236525-MS) built an agent to find invisible lost time (ILT), the gap between actual performance and the technical limit, where reconciling rig sensors, activity codes and the plan had stayed manual. ML classifiers set rig states from high-frequency sensor data and LLM agents reconcile them into one drilling timeline, with provenance on every interval.


The whole system runs inside an air-gapped operator tenant, the low-frequency report stays the auditable record, and every published figure passes a human-review gate. The pilot covered five wells and two rigs in an offshore field during 2026; closed-loop optimization is left explicitly to future work. The authors also record where the agent over-flagged, where a human reviewer overrode it and where it added nothing.


🎯 Your move: for every agent you pilot, record how often a reviewer overrides it and how long the review takes; that is the cost the speed figures leave out (WillCo analysis).



📏 Measurement: Sensing at the Bit, Decoding the Pulse

Two SPE Journal Papers Push Measurement Closer to the Cutter and Cleaner Through the Mud

SPE Journal papers SPE-236979-PA and SPE-237003-PA, November 2026, with Crossref records deposited November 5 and November 9.


Two peer-reviewed papers target downhole measurement: one senses force at the cutter itself, tested in a cutting experiment, and one recovers weak mud-pulse telemetry from noise on data from three wells.


Sensing at the cutter (SPE-236979-PA): a PDC cutter aimed at online cutter health evaluation, lithology identification and closed-loop control of intelligent drilling systems. The SCutter senses its internal stress and, through whole-cutter calibration, estimates the cutting force acting on it, with two temperature sensors on board; in a cutting experiment, its response tracked an external triaxial force sensor with a correlation of 0.943 (Pearson r). The authors call it a first of its kind and report a mean relative tracking error of 8.39% across five cutting depths. They are from Southwest Petroleum University and the University of Aberdeen with partners.


Cleaner mud-pulse telemetry (SPE-237003-PA): IVBAKS, an improved Kalman-smoother method, recovered weak mud-pulse signals from noise on surface-acquired data from three wells, with a bit error rate (BER) of at most 0.02 on the evaluated standard sequence segments. In simulation at a -15 dB signal-to-noise ratio, its average BER was 0.08, and combined with standard Kalman suppression of pump noise, the scheme recovered the pulse waveforms in ground hydraulic circulation tests. The authors are from China University of Petroleum (East China) and industry partners.


🎯 Your move: list which of your loop's inputs are measured at the bit and which are inferred from surface data or decoded from mud pulses, and how often each one arrives; a cleaner decode means more usable data per hour from the same pulser (WillCo analysis).



🧩 Standards: The Vocabulary Has Gaps

Two Decades of Well-Intervention Records Meet OSDU, and Well-Control Pressures Do Not Fit

Published November 2 in the ADIPEC 2026 proceedings (SPE-235838-MS).


Well-control pressures and HSSE observation counts have no direct home in OSDU R3, an AEM Energy Solutions and PETRONAS Digital proof of concept finds.


The problem is scale and age: about 2,000 wells and two decades of operations, with records spread across five to ten legacy repositories, each with its own vocabulary, structure and vintage. The authors propose a five-service pipeline aligned with TOGAF 10: Source Inventory, Extraction, Ontology Mapping, Transformation, and Load and Governance.


A rule-based proof of concept mapped real WIDR records onto OSDU R3 schema kinds using a deterministic dictionary. 🚨 The principal finding is the gap: a substantial share of well-intervention fields, notably well-control pressures and HSSE observation counts, have no direct home in the standard.


🎯 Your move: before any migration, list the fields your own reports depend on that the target standard cannot hold, and decide who owns the extension (WillCo analysis).



📋 In Brief

Each item ends with a question to put to your own team (WillCo analysis).


Drilling performance

  • ⚡ Patterson-UTI: a reinforcement-learning generator advisor becomes the default: according to Patterson-UTI's abstract for the November 10 IADC DEC forum, its reinforcement-learning (RL) generator-management advisor was trained on data from 104 land rigs and, after fleet-wide testing, became the contractor's default system. The abstract says it reduced average generators online by more than 25% versus average fleet performance, cut generator-cycling events by 58% and reduced recommendations risking power-limit events by 87%. Autonomous control remains a future step. Read: the clearest item in the window to move from available to default use, and it is generator-management advisory, not autonomous wellbore control (WillCo analysis). Which of your automated advisors is the default today, and who decided it?

  • 🚀 OQEP and SLB: on-bottom ROP up 56% in the 8-1/2-in. section: SLB and OQEP authors report section-level results of autonomous and advisory drilling from three trial wells against field offset wells: on-bottom ROP up 17% in the 12-1/4-in., 56% in the 8-1/2-in. and 12% in the 6-1/8-in. section. (Follow-up to Edition 7's SLB Neuro item; SPE-236416-MS, November 2.) Which of your sections would gain most from closing the loop first?

  • 🔁 PDO MSE digital twin: drilling time down up to 17%, well cost about 6%: PDO's author scores the model on agreement between its recommended action and the recorded outcome of each bit run, validated on data from more than 500 wells across 15 fields; in one case study the reservoir section took three bit runs against eight in an offset well. (SPE-235993-MS.) Can your real-time model show how often its advice matched what happened?

  • 🧭 Sirt Oil and SLB: 50% fewer downlinks in a thin Libyan lateral: five Sirt Oil and 14 SLB authors report that a fully autonomous geosteering workflow kept the wellbore within the target interval and within 5 to 7 ft of the reservoir top. (SPE-235846-MS.) How many downlinks per lateral does your steering need today?

  • ⚙️ Weatherford MPD: bottomhole pressure held with pumps off, zero liner losses: two authors from Weatherford's MPD business report a field application that held the target bottomhole pressure with the pumps off and ran and cemented the 7-in. liner with zero losses. Against a deliberately lightened static mud column, it applied on the order of 1,090 psi of surface backpressure and drilled the section 1,306 ft to TD on a single bit. (SPE-235533-MS.) Does your MPD control logic cover the moments when you cannot circulate?


Operations and data

  • 🛰️ SLB coiled tubing from a remote center: pipe failures eliminated in the first full year: SLB authors report on their own remote operations center framework, validated in more than 100 jobs, and say pipe failures were eliminated entirely in the first full deployment year. (SPE-236316-MS.) Which of your intervention fleets could be supervised from one room?

  • 🔦 ADNOC Onshore and Weatherford: fiber-optic integrity interventions from days to hours: a dedicated capillary unit recorded simultaneous DTS and DAS in a single trip per well, and the authors report intervention duration moving from days to hours. (SPE-235991-MS.) How long does your next well-integrity diagnosis take, from first signal to answer?

  • 🧹 CNPC Chuanqing: usable fluid records from 77.0% to 98.0%: before cleaning, 77.0% of the online fluid-monitoring records in a replay dataset were usable; after cleaning, 98.0%. A simplified cleaning operator that keeps operational changes recovered all 55 genuine events, at a 25.61% false-alarm rate, where a three-sigma rule caught none. (Sensors, November 5.) What share of your online fluid records would pass a quality check today?

  • 🤖 Weatherford: AI coding agents shorten field-to-workflow from months to days: two Weatherford authors report that a domain engineer directing general-purpose AI coding agents built a first usable fiber-sensing interpretation environment within weeks, and say the approach shortens the cycle from field observation to reproducible workflow from many months to a few days. (SPE-235234-MS.) Who validates the code an agent writes before it touches well data?

  • 🎓 SLB authors: a skills engine for an AI-ready workforce: SLB authors report a skills engine deployed across about 8,000 professionals at an energy organization; it delivers curated curricula for aspirational roles, including promotion-linked ones. (SPE-235223-MS.) Who on your team is trained to supervise automation, not just run it?


Market and vendors

  • 📈 Pason: record revenue per industry day, up 5%: Pason reports a record North American Drilling Revenue per Industry Day of C$1,125 in the third quarter, up 5% from C$1,071 a year earlier, driven by product adoption and Canadian activity. Pason names two strong tailwinds: industry adoption of data-driven automation and analytics, which relies on reliable, high-quality data, and rising well complexity. (Third-quarter results, November 5.) Is your data contract priced on rig days or on the data quality your automation needs?

  • 🗺️ SLB and Crescent Petroleum: three-year digital field development planning contract: SLB calls it the first commercial adoption of FDPlan in MENA. Crescent's existing Delfi use, not FDPlan, cut simulation times on selected large dual-porosity, dual-permeability models from about six hours to around 50 minutes and lifted runs from about 20 to about 140 in one month. (SLB newsroom, November 5.) How many simulation runs does your plan get before spud?

  • 🏭 AIQ and Siemens Energy: RoboWell autonomous well control among the first offerings: the agreement, announced at ADIPEC (November 2-5), names AIQ's RoboWell autonomous well control application among the first offerings; the release cites an ADIPEC demonstration and measurable results in the UAE but names no joint customer deployment. (Zawya, November 9.) What evidence would RoboWell have to show before you let it control one of your wells?

  • 🌐 Shenkai claims 15% shorter average drilling cycles across 30+ units: Shenkai brought its full-chain AI wellsite to ADIPEC and says its products have been applied at more than 30 oilfield units, with the average drilling cycle shortened by 15% and an average reservoir encounter rate above 90%. Shenkai also says it agreed with the ADNOC Drilling CEO to advance deployment, a prospective step. (Company WeChat post, November 4, in Chinese.) Which of these vendor figures would your team need to see on its own wells?

  • 🧱 Seequent buys Eliis: PaleoScan seismic interpretation moves into a subsurface modeling platform: the deal brings Eliis's PaleoScan into Seequent, which says the deal widens the subsurface modeling it offers to geothermal, oil and gas, and carbon capture. (ThinkGeoEnergy, November 4.) Can you trace which seismic interpretation your geosteering is reading?


📡 On WillCo's Radar



Why it matters: a service company now measures automation in rig time saved per month and in reliability, not in tasks automated: the same move the Top Story makes from installs to coverage (WillCo analysis).

  • 🎙️ SLB's Rakesh Jaggi on the pace of autonomy (Energy Connects, November 4): at an ADIPEC panel, the consensus favored human-augmented AI, with people setting intent, context and guardrails. Jaggi, President of Digital at SLB, invited the audience to "dream a little": with connectivity, edge AI and robotics combined, he said drilling with no humans present is possible today, but that being 99.99% sure is not good enough and the transition may be a little on the slow side. Read: the technology question is settled faster than the trust question (WillCo analysis).

  • 🧪 Red Hat and Amrita on bounding AI agents (SPE-236401-MS and SPE-235393-MS): both tested on a simulated chemical reactor, not a rig. Bounded autonomy: budget limits cut safety-check rejections for reactive agents by 89%, and cross-model tests on two other models confirmed the direction of the violation-reduction effect at a smaller magnitude. Shadow plant: a propose-validate-correct layer intercepted proposals that would have violated safety constraints, with zero actual violations in all 282 cells, and the authors say it does not replace a certified SIS. Read: limits on what an agent may do are becoming an engineering design input (WillCo analysis).

  • 🔊 Baker Hughes tests a lightweight edge AI agent for DAS anomaly detection on synthetic and field DAS datasets (SPE-235621-MS): Baker Hughes authors report that an unsupervised agent with signed agent manifests, execution-path control, a watchdog and fail-safe modes cut communication by 94.36%, with average latency below one second and no accelerator hardware. Read: trust features are being built into the agent itself, not added afterward (WillCo analysis).


🔭 What We're Watching Next

  • 🦾 H&P and Bentec robotics results (IADC DEC Tech Forum, Houston, November 10): H&P and Bentec's abstract for the forum framed the first fully robotic connection, the first hands-free tripping sequence and the first multi-pad deployment of H&P's FlexRobotics program as its 2026 firsts. The abstract cites field data from thousands of robotic cycles and reliability metrics; the public program does not publish the metric values. Those values are the evidence to watch (WillCo analysis).

  • 🛰️ Corva's year of firsts (IADC DEC Tech Forum, Houston, November 10): Corva's abstract highlighted four 2026 firsts for its Predictive Drilling: use with four new drilling contractors, an extended-reach offshore lateral in Advisory Mode, a U-Turn well and programmable stand-by-stand recipes. The public program carries no case-level result figures; the results behind each first are what to watch.

  • 🌎 SPE Workshop on Digital Transformation, Rio de Janeiro, November 10 to 11: the workshop ends November 11; its drilling and well intervention session program lists Halliburton on remote, autonomous deepwater cementing, Petrobras on moving from digital-twin monitoring to smart-twin-based automation, and ROGII's talk "A Million Wells and AI: Bringing Order to the Factory".

  • 📝 IADC ART Committee, November 11: the committee's next meeting is listed for November 11. Its August agenda carried a report on the AI landscape in drilling, the IADC/DWIS Lexicon and a rig automation survey.



When does automation stop being a pilot and become the way a rig drills? When someone measures how much of the work it actually does.


Three stages: available, routinely used, default. This week's papers sit on the first two; one contractor abstract claims the third.


Availability: installed and working, as automated steering was on the Oman rigs.


Routine use: crews let it run most of a section, as on the Field A well where DPO ran 81% of the section; in this well set, the high-coverage wells were the fast ones.


Default: the automated method is how the work is done unless someone decides otherwise. In its IADC DEC abstract, Patterson-UTI says its reinforcement-learning generator-management advisor became the contractor's default system after fleet-wide testing; closed-loop drilling optimization in the wellbore remains earlier on the path.


The tension is verification. Coverage rises only when crews trust the output, and trust has a cost: the ADNOC Offshore pilot keeps a human-review gate on every figure, Expro's COO describes keeping override authority with qualified people, and the Red Hat papers build limits around the agent, checked before any action reaches the process. Read that as design, not caution: it is what lets coverage grow without a failure that sends it back to zero.


Underneath the path sit the inputs and the vocabulary: a loop cannot run more of the section than its measurements can see, and fleet-scale defaults need records that mean the same thing everywhere.


Automation becomes the default the same way a new procedure does: it is used, measured and trusted on ordinary wells until no one asks to turn it off. Operators who measure coverage and overrides will know when they have reached that point; those who count installations will not.


The contract test: put coverage, overrides and hand-back conditions into the next performance contract (see Practical Recommendations).


✅ Practical Recommendations

For operators and drilling contractors: treat coverage as a performance metric, not a feature list, and write three terms into the next performance contract: coverage, override rate and the conditions under which the system hands back control. (WillCo analysis)

  1. 📊 Measure coverage this month: for each well, record the share of each section the optimization or automation system actually ran, by rig and by crew.

  2. 🔍 Ask why low coverage was low: sort low-coverage sections by cause: crew choice, equipment limits, or a workflow that stops at the hard parts.

  3. ⚖️ Count overrides on every agent pilot: log how often a reviewer corrects or overrides the agent, and how long each review takes, and judge the pilot on time to a signed answer.

  4. 📏 Map your loop's inputs: list which inputs are measured downhole, which are inferred from surface data and which are decoded from telemetry, with their update rates.

  5. 🧩 List your standards gaps before migrating: name the fields your reports depend on that the target standard cannot hold, and assign an owner for each extension.


🛠️ Start with one rig: score it on the free 15-question Closed-Loop Readiness scorecard.


📚 References and Sources

  1. Marwa Al Sulaimani, Emad Diab, Anton Zaytsev, Ahmed Sami Allam, Ali Al Maimani, Ahmed Nabeel, Samya Al Sawafi, Salah Habachi, Hussein Al Kiyumi, Bader Al Qanubi (Halliburton); Mohamed Al Hinai (Petroleum Development Oman). "Enhancing Well Delivery Through Integrated Drilling Automation and Real-Time Geosteering in Upper Shuaiba Reservoirs," SPE-235814-MS, ADIPEC 2026. DOI: 10.2118/235814-MS. https://doi.org/10.2118/235814-MS

  2. Z. Mushtaq, A. O. Alshaigy, B. Kodiah, M. Husien, F. Alhosani, B. Banihammad, H. Almarshoodi, S. Rozyyev, D. Shaikh (ADNOC Onshore); A. Galimkhanov (Corva). "AI-Driven Predictive Modeling for Real-Time Drilling Optimization in Extended-Reach Wells: Enabling First Trend-Setting Mile-A-Day Performance for a UAE Operator," SPE-235580-MS, ADIPEC 2026. DOI: 10.2118/235580-MS. https://doi.org/10.2118/235580-MS

  3. F. Landaeta Rivas, M. Khan, C. Andre Coudyzer, K. Alghaithi, A. I. Handoko, Y. Zhao, A. N. Awwad, L. Toader, S. M. Al Ameri, S. Alblooshi, Y. Al Katheeri (ADNOC Offshore); A. Khoshnaw (Corva Ai); A. Galimkhanov (Corva). "Agentic AI in the Air-Gap: Hybrid ML and LLM Reconciliation of Combined-Frequency Drilling KPIs at ADNOC Offshore," SPE-236525-MS, ADIPEC 2026. DOI: 10.2118/236525-MS. https://doi.org/10.2118/236525-MS

  4. P. Sesboüé, D. F. Patino, O. Bustos Angel, V. Guillot, S. Sebbane (SLB). "From Tender to Total Depth: Artificial Intelligence and Generative AI Agents Deployed Across Well Construction Planning and Execution," SPE-236279-MS, ADIPEC 2026. DOI: 10.2118/236279-MS. https://doi.org/10.2118/236279-MS

  5. Wenwei Wang, Can Cai et al. (Southwest Petroleum University; University of Aberdeen; Nanchong Key Laboratory of Robotics Engineering and Intelligent Manufacturing; Missan Oil Training Institute, Ministry of Oil, Iraq). "Development of a Smart Polycrystalline Diamond Compact Cutter Featuring In-Situ Temperature and Stress Sensing," SPE-236979-PA, SPE Journal, 2026. DOI: 10.2118/236979-PA. https://doi.org/10.2118/236979-pa

  6. Xueyang Zhao, Zhidan Yan, Ruirui Sun et al. (China University of Petroleum (East China) and industry partners). "Improved Variational Bayesian Adaptive Kalman Smoother Method for Random Noise Suppression in Continuous-Wave Mud-Pulse Telemetry," SPE-237003-PA, SPE Journal, 2026. DOI: 10.2118/237003-PA. https://doi.org/10.2118/237003-pa

  7. M. K. Ikhwan (AEM Energy Solutions); N. S. Che Zakaria (PETRONAS Digital). "Legacy to OSDU: A Software Engineering Blueprint for Migrating Siloed Well Intervention Data to Open Standards," SPE-235838-MS, ADIPEC 2026. DOI: 10.2118/235838-MS. https://doi.org/10.2118/235838-MS

  8. "SLB to Deploy Digital Field Development Planning for Crescent Petroleum" - SLB newsroom, Nov 5, 2026. https://www.slb.com/newsroom/updates/2026/slb-to-deploy-digital-field-development-planning-for-crescent-petroleum

  9. "Pason Reports Third Quarter 2026 Results and Declares Quarterly Dividend" - Pason Systems news release via Canada NewsWire (Placera.se mirror), Nov 5, 2026. https://www.placera.se/pressmeddelanden/pason-systems-pason-reports-third-quarter-2026-results-and-declares-quarterly-dividend-20261105

  10. Qi Lu, Zhilin Li, Dongxiao Pang (CNPC Chuanqing Drilling Engineering). "Evaluation and Remediation of Data Quality for Drilling-Fluid Property Online Monitoring," Sensors, Nov 5, 2026. DOI: 10.3390/s26217009. https://doi.org/10.3390/s26217009

  11. M. Hammad et al. (Sirt Oil Company; SLB). "Autonomous Geosteering Workflow for High-Precision Well Placement and ROP Optimization: A Thin-Reservoir Case from Libya," SPE-235846-MS, ADIPEC 2026. DOI: 10.2118/235846-MS. https://doi.org/10.2118/235846-MS

  12. H. A. Humaidi et al. (SLB; OQEP). "Implementation and Results of Neuro Autonomous Solutions and DrillOps Advisory in OQEP Onshore Wells," SPE-236416-MS, ADIPEC 2026. DOI: 10.2118/236416-MS. https://doi.org/10.2118/236416-MS

  13. Abdullah Al Mamari (Petroleum Development Oman). "Real-Time Digital Twin Model for Optimizing Drilling Performance and Support Intelligent Decision Making," SPE-235993-MS, ADIPEC 2026. DOI: 10.2118/235993-MS. https://doi.org/10.2118/235993-MS

  14. Ahmed Al Dahi, Carlos Iturrios (Weatherford). "Extending the Operating Envelope: Integrating Constant Bottomhole Pressure Managed Pressure Drilling to Maintain CBHP When Internal Circulation is Restricted," SPE-235533-MS, ADIPEC 2026. DOI: 10.2118/235533-MS. https://doi.org/10.2118/235533-MS

  15. William Tapia, Amir Naveed, Gleidson Alves (SLB). "Digital Revolution in Coiled Tubing Operations: Pioneering a Remote Operations Center Framework that Redefines Intervention Efficiency, Asset Reliability, and Sustainability Across a Large-Scale CT Fleet," SPE-236316-MS, ADIPEC 2026. DOI: 10.2118/236316-MS. https://doi.org/10.2118/236316-MS

  16. Arjun Guglani, Ehsan Nikjoo (Weatherford). "AI-Agent-Built Fiber-Sensing Interpretation Platform for Real Well Operations," SPE-235234-MS, ADIPEC 2026. DOI: 10.2118/235234-MS. https://doi.org/10.2118/235234-MS

  17. "Shenkai at ADIPEC 2026: full-chain AI smart wellsite capability" (in Chinese) - Shenkai Petroleum WeChat via 10jqka, Nov 4, 2026. https://news.10jqka.com.cn/20261104/c681447000.shtml

  18. "AIQ and Siemens Energy announce collaboration to advance autonomous digital capabilities for energy and process industries" - AIQ via Zawya, Nov 9, 2026. https://www.zawya.com/en/press-release/aiq-and-siemens-energy-announce-collaboration-to-advance-autonomous-digital-capabilities-for-energy-and-process-industries-2127040

  19. "Seequent acquires seismic interpretations software company Eliis" - ThinkGeoEnergy, Nov 4, 2026. https://www.thinkgeoenergy.com/seequent-acquires-seismic-interpretations-software-company-eliis/

  20. A.D. Carolina et al. (SLB). "The Skills Intelligence Engine for an AI-Ready Energy Workforce," SPE-235223-MS, ADIPEC 2026. DOI: 10.2118/235223-MS. https://doi.org/10.2118/235223-MS

  21. Manish K. Sinha et al. (ADNOC Onshore; Weatherford). "From Days to Hours, from Inference to Evidence: Fiber-Optic Sensing as an Efficient, High-Fidelity Well Integrity Diagnostic Tool: ADNOC Onshore's First Trial Workflow," SPE-235991-MS, ADIPEC 2026. DOI: 10.2118/235991-MS. https://doi.org/10.2118/235991-MS

  22. A. Hurtado. "ADIPEC 2026: Fully autonomous rigs remain a distant prospect" - Offshore Magazine, Nov 3, 2026. https://www.offshore-mag.com/equipment-engineering/article/55416176/expro-adipec-2026-fully-autonomous-rigs-remain-a-distant-prospect

  23. "Energy leaders see major opportunities in AI, with autonomous drilling in sight" - Energy Connects, Nov 4, 2026. https://www.energyconnects.com/opinion/features/2026/november/energy-leaders-see-major-opportunities-in-ai-with-autonomous-drilling-in-sight/

  24. M. I. Khan, M. Khanna et al. (Red Hat; Amrita Vishwa Vidyapeetham). "Bounded Autonomy: A Budget-Constrained Framework for AI Agent Operations Toward Safety-Critical Process Environments," SPE-236401-MS, ADIPEC 2026. DOI: 10.2118/236401-MS. https://doi.org/10.2118/236401-MS

  25. M. I. Khan, M. Khanna et al. (Red Hat; Amrita Vishwa Vidyapeetham). "Propose-Validate-Correct Shadow-Plant Architecture for Safe AI Agents in Process Operations," SPE-235393-MS, ADIPEC 2026. DOI: 10.2118/235393-MS. https://doi.org/10.2118/235393-MS

  26. P. Moradi, N. Kalia, J. Noguera (Baker Hughes). "Lightweight Edge AI Agent for Real-Time Anomaly Detection in Distributed Acoustic Sensing," SPE-235621-MS, ADIPEC 2026. DOI: 10.2118/235621-MS. https://doi.org/10.2118/235621-MS

  27. "IADC DEC Tech Forum Q4" - IADC event page, Nov 10, 2026. https://iadc.org/event/iadc-dec-tech-forum-q4/

  28. "Session 3: Application of Advanced Digital Technologies to Drilling and Well intervention" - SPE Workshop: Digital Transformation, Rio de Janeiro, Nov 10 to 11, 2026. https://www.spe-events.org/workshop/digital-transformation/technical-program/session-3-drilling-well-intervention

  29. "IADC ART Committee" - IADC event page for the August 18, 2026 meeting, listing the next meeting on Nov 11, 2026. https://iadc.org/event/iadc-art-committee-3/

  30. Prior WillCo Drilling Intelligence editions and archive - https://willcodrilling.com/blog


About this report: The WillCo Drilling Intelligence Report is a weekly synthesis of verified developments in digital drilling and well construction, published by WillCo Drilling Consulting. Analysis, forecasts, and recommendations reflect WillCo's independent engineering perspective.



📖 Glossary of Acronyms



Comments


bottom of page