Frequently Asked Questions
To provide clarity around how AI-driven production optimization works in practice, we’ve compiled answers to the questions we hear most often. This FAQ brings together common operational, technical, and lift-specific questions to offer a clear view of how the platform is used across different production environments.
Q1. How do you handle downhole anomalies in the field?
A: Typically a loss in production is the first indicator of a downhole anomaly. This can occur gradually over time - or happen immediately. Next is diagnosis, identifying potential root causes of the DH anomaly, and then finally generating a solution to resolve or mitigate the downhole anomaly.
Q2. What issues can Ambyint detect but not directly resolve?
A: Ambyint can quickly identify issues through our anomaly detection suite and all anomalies are indicators that either the sensors associated are compromised, or an operational event is occurring which cannot be resolved through setpoint management. For example, in plunger lift operations this may present as a leaky valve anomaly, while in rod lift systems it could appear as a worn pump anomaly. In both cases, these issues restrict production and require mechanical intervention.
Q3. What operational changes can Ambyint make to wells?
A: Ambyint’s Automated Setpoint Management (ASPM) can make setpoint adjustments with the purpose to increase production. For plunger lift, ASPM can manage open and close triggers - directly affecting cycle characteristics. For rod lift, ASPM can speed up or down SPM, and also adjust idle time for best individual well performance.
Q4. What safety limitations govern autonomous adjustments?
A: Ambyint ASPM is bound by customer defined safe operating parameters. Additionally, ASMP has internal backstops to limit the rate of change, up to a max 10% of current setpoint values to ensure that large setpoint step changes do not occur.
Q5. How are wells tagged or organized by routes, regions, or foremen?
A: Ambyint provides the customer the ability to upload their own well hierarchy, including routes, areas, foreman, operators, engineers. The Engineering and Operations teams are able to quickly focus only on those wells they are responsible for. In addition Ambyint provides a unique tagging feature that gives the user a shortcut to the wells they want to see. Ambyint’s Tagging system is two fold with system generated tags and user created tags.
Q6. Can the platform be used without AI functionality enabled?
A: Absolutely! The Infinity Suite has the ability to generate recommendations that can be reviewed and if the user chooses to, update the setpoint through the Ambyint platform. ASPM can be enabled at any time if the user chooses to automatically accept future recommendations. With Ambyint’s unique analytics, users can also have all their wells in the platform without AI recommendations enabled, taking advantage of the analytical data generated by lift specific for additional insights into their well performance.
Q7. How often can Ambyint receive or ingest data from the controller?
A: Ambyint can accept a wide range of data quality. From 1 min data to 30 min data quality, the Infinity solutions will still generate high quality recommendations for setpoints to optimize the well. Data resolution of greater than ~20 minutes can impact reliability and accuracy of some anomaly detection functionality, as many of the data signatures the anomaly detection models are looking for can be missed between data points.
Q8. How do operators validate or review Ambyint’s recommendations?
A: Each well will generate manual setpoint recommendations when all parameters are met. The manual recommendations can be reviewed per well, and if the operator agrees with the recommendations, implemented directly from the Ambyint platform to the well RTU by clicking update. Once the operator is comfortable with the manual recommendations, they can choose to enable ASPM, which will allow the recommendations to automatically be implemented to the well when generated.
Q9. What visibility do field teams have into autonomous actions?
A: We have a full event log available for the users! We log all events for the well setpoints managed through the platform including: mode changes, manual recommendation generation, implementation, ASPM setpoint requests and changes. We also keep the event history of the well forever, so any changes that occurred on the well, either from Ambyint or from a SCADA setpoint change, is documented and logged for all to review, for all time.
Q1. What does the data integration process look like, and how is data transferred from SCADA into Ambyint?
A: Ambyint’s Implementation and Integration teams work collaboratively with the customer’s IT and SCADA teams to ensure safe, secure, and seamless two-way communication between Ambyint’s cloud-based platform and the customer’s SCADA/OT systems. All or most of the required data lives in a customer’s SCADA system, so this is the focus for data integration. In most cases, Ambyint is able to complete the data integration process and fully stand-up the solution at scale within eight weeks.
We leverage secure MQTT protocols to enable safe and secure two-way communication with a customer’s SCADA system. Ambyint has been SOC2 Type 2 compliant for 4 years, with SOC2 audits and penetration tests conducted on an annual basis.
Q2. What IT or network requirements are needed for onboarding?
A: The customer must have the ability to connect to Ambyint’s MQTT broker to publish and subscribe to data. Many customers already have this in place, making integration relatively seamless. For customers that do not have this ability in place today, we work with 3rd party SCADA integrators and the customers to quickly and cost-effectively establish the capability. Ambyint does NOT require any inbound firewalls to be opened.
Q3. What are the phases of a standard pilot rollout?
A:
1. Planning and preparation - data review and timeline confirmations
2. Integration - well set up and configuration of SCADA and Ambyint platform
3. Validation - data review and write back confirmation
4. Execution - well reviews, move wells to autonomous mode
5. Pilot Review
Q4. What validation steps occur before autonomy is enabled
A: Data is reviewed and validated with customers, and then set points recommendations are reviewed and validated with a focus on potential impact to safety, productivity, and reliability. Wells are not put into autonomous mode without confirmation from the customer.
Q5. Why do operators trust Ambyint’s ASPM?
A: Trust is earned and not given. As part of the validation process, Ambyint SME’s work closely with our customers to review, discuss and manually implement setpoint recommendations generated in our platform. As setpoint recommendations are generated every 48 hrs, Ambyint is committed to earning our customers trust through this process until they are completely satisfied and comfortable with the set point recommendations to be switched to full auto implementation.
Q6. How are actions, recommendations, and changes tracked for auditing?
A: Recommendations and actions taken by the platform are displayed in the events panel of the platform. Each transaction is also recorded in Ambyint’s logs.
Q7. Who can I contact if I experience issues with the platform?
A: You can click the question icon in the lower-left corner of the platform for in-app support, or email support@ambyint.com, our Customer Support or Customer Success teams will assist you.
Q1. What types of plunger lift anomalies can Ambyint detect?
A: Leveraging true machine learning and our advanced physics engine, Ambyint can detect a range of plunger lift anomalies that impact performance. Whether it’s the wellbore, on the surface, transmitters or transducers, flow meters, or the plunger itself, Ambyint’s anomaly detection will flag these issues before they become more costly problems.
Q2. Where can I see a full list of anomalies and descriptions or each anomaly?
A: All currently supported plunger anomalies are displayed in the Ambyint platform. Models are always being trained to both expand to new anomalies, as well as continue to polish active and current ones. A full list of supported plunger lift anomalies, along with detailed descriptions and example data, is available in our Plunger Lift Field Guide. Our Plunger Lift Field Guide has a full list of plunger lift anomalies, along with detailed descriptions on the anomalies and what to look for, useful even as a cheat sheet!
Q3. How does the platform assess well readiness for a plunger cycle?
A: The Ambyint platform assesses well readiness with a variety of checks and balances by analyzing key operational signals. This includes observing realistic pressure and rate data, as well as plunger arrivals. With all this data in hand, the platform must also successfully analyze these complete cycles and generate recommendations as an additional check prior to enabling autonomous setpoint management (ASPM).
Q4. What data sources are used to calculate plunger diagnostics?
A: For plunger diagnostics, Ambyint will source pressures, rate, and arrivals from your SCADA system. With our data agnostic approach we can pull the necessary data from any system to ensure support for customers with all architecture types.
Q5. How does Ambyint improve cycle time or optimize shutdowns?
A: Ambyint improves cycle time by identifying suboptimal on and off times,excessive shut-in durations, or premature openings that reduce lift effectiveness. By analyzing historical cycle outcomes and real-time conditions, the platform highlights these opportunities and avoids failed or marginal cycles.
Q6. Can Ambyint recommend or automate plunger cycle adjustments?
A: Both. Although the ultimate goal is to enable ASPM and minimize human interaction, we always start by pushing recommendations that must first be approved by a human as to let our customers build trust in the system. Manual setpoint changes initially are also recommended as to let Ambyint's models calibrate to your unique assets.
Q7. What communication frequency is necessary for accurate plunger tracking?
A: Ambyint can successfully automate and optimize plunger lift wells with a range of different polling frequencies, although finer resolution data will improve the effectiveness of the system - most notably with anomaly detection. The minimum data resolution for successful plunger optimization is 15 min sampling of pressures. Polling can be less frequent as long as the data is sampled every 15 min.
Q8. How does Ambyint handle gaps in cycle data or missed arrivals?
A: Our platform is designed to handle outages with validation, interpolation, and contextual analysis. The Ambyint platform will also flag periods of missing data when optimization and automation is inhibited. The Ambyint platform serves as a layer that sits on top of existing real-time automation equipment in the field. Ambyint manages the setpoints that govern how this equipment operated, but ultimately, even if there is a data outage, the automation systems in the field will continue to function normally and safely.
Q9. How can field teams validate plunger performance trends in the platform?
A: Ambyint's surveillance and analytics tools are designed so that our customers can quickly drill down to the wells needing attention. The customizable reporting feature also ensures the correct data gets into the correct hands at a desired frequency, so that the highest priority action items can be tackled quickly.
Q10. What historical data is stored and for how long?
A: 18 months of data is stored. This includes your wellhead data, pressures, rates, and cycle data.
Q11. What happens if my well goes offline?
A: If a well goes offline or stops communicating, the platform flags the lost communication and preserves the most recent valid data. Once communication resumes, the system automatically reconciles new data, if it's available, with historical context to restore diagnostics, tracking and autonomous control.
Q12. How long after implementation will I start seeing results?
A: Many customers begin seeing actionable setpoint recommendations within days of connecting to the platform. Early results typically include improved visibility, faster anomaly identification, and reduced time spent manually reviewing wells. Once the platform is calibrated to your assets and recommendations are verified, full ASPM can be enabled in as little as two weeks.
Q13. How long after implementation will the ASPM start working?
A: Once the platform has been calibrated to your unique assets and sufficient baseline data is collected, ASPM can start to be switched on. This could happen in as little as two weeks with correct initial oversight.
Q1. What types of rod lift anomalies can Ambyint detect?
A: Ambyint can detect a wide range of rod lift anomalies using a combination of physics-based diagnostics and machine-learning classification. Common conditions include fluid pound, gas interference, pump fillage loss, worn pumps, tagging, sticking pumps, rod overloading, downhole friction changes, and surface equipment issues. The system continuously evaluates well behavior to identify both acute failures and slowly developing degradation trends.
Q2. Where can I see a full list of anomalies and descriptions for each anomaly?
A: A full list of supported rod lift anomalies, along with detailed descriptions and example dynacards, is available on our Rod Lift Field Guide. Our Rod Lift Field Guide has a full list of rod lift anomalies, along with detailed descriptions on the anomalies and what to look for, useful even as a cheat sheet!
Q3. What data sources are used to calculate rod lift diagnostics?
A: Ambyint leverages the data generated by a conventional pump off controller, including surface and downhole dynacards, process control time series data, and setpoint data. In addition, we incorporate rod design, pump details, and wellbore information to accurately configure our physics-based models.
Q4. What communication frequency is required for accurate rod lift tracking?
A: While higher frequency data improves resolution and responsiveness, the platform can still deliver accurate diagnostics and insights with lower-frequency polling common in many rod lift operations. Typically, we ask for 15 min data.
Q5. How does Ambyint handle gaps in data?
A: The platform is designed to handle outages with validation, interpolation, and contextual analysis. The Ambyint platform will also flag periods of missing data when optimization and automation is inhibited. The Ambyint platform serves as a layer that sits on top of existing real-time automation equipment in the field. Ambyint manages the setpoints that govern how this equipment operated, but ultimately, even if there is a data outage, the automation systems in the field will continue to function normally and safely. Pumpoff controllers typically historicize much of the data, so as soon as communications are restored, we can continue to make changes at the well.
Q6. How are rod string details incorporated into diagnostics?
A: Rod string information such as rod sizes, grades, taper design, and pump depth is used to enhance load interpretation and dynacard analysis. When complete well configuration data is available, Ambyint applies this information to improve anomaly classification and reduce false positives. For computational simplicity, most conventional POCs do not factor in wellbore deviation and key friction forces, resulting in lower accuracy downhole cards. Ambyint reprocesses all surface cards through our proprietary deviated wellbore downhole wave equation that has been deployed commercially for nearly 20 years.
Q7. What is a dynacard?
A: A dynacard is a plot of load versus position for a rod pump during a pumping cycle. It provides insight into downhole conditions and pump performance by visualizing how forces change throughout the stroke, making it one of the most valuable diagnostic tools in rod lift operations.
Q8. How does Ambyint use dynacards?
A: Ambyint analyzes dynacards at scale using automated classification and pattern recognition. Instead of requiring manual review, the platform continuously evaluates card shapes to identify operating conditions, detect anomalies, and track changes over time across entire well populations.
Q9. How does Ambyint leverage historical cards in anomaly detection?
A: Historical dynacards are used to establish a baseline of normal behavior for each well. By comparing current cards to historical patterns, Ambyint can detect subtle deviations, identify early signs of failure, and differentiate between transient issues and persistent problems.
Q10. How can operators compare historical and current card behaviour in the platform?
A: Operators can easily visualize and compare historical and current dynacards directly in the Ambyint interface. Custom views, trend plots, and time-based filtering allow teams to quickly assess how a well’s behavior has evolved and validate the impact of operational changes.
Q11. What historical data is stored and for how long?
A: 18 months of data is stored on the platform
Q12. How long after implementation will I start seeing results?
A: Most operators begin seeing actionable insights within days of onboarding. Early results typically include improved visibility, faster anomaly identification, and reduced time spent manually reviewing wells. It’s not uncommon for us to see 15-20x ROI during the course of a pilot.
Q13. How long after implementation will the ASPM start working?
A: ASPM can begin operating within 2 weeks after implementation once sufficient baseline data is established and control parameters are validated. In many cases, automated setpoint adjustments and optimization actions can start within weeks, depending on data quality, operator preferences, and operational readiness.
Have more questions?
Get in touch with the team today.
