I was talking with a partner this week, and I of course asked him about the deals he is working where Hedgehog may be a fit. He mentioned an energy company in Houston, and I had a flashback to 2005.
My first customer that year was a company in Houston called Universal Compression (now Enerflex). UCI manufactured natural gas compressors that moves gas through pipelines. They priced their contracts with service level agreements that guaranteed a certain throughput, so measuring and maintaining gas pressure was a critical operational requirement. UCI had already invested in telemetry in their products well before the term “Internet of Things” was invented. Sensors on the compressors measured pressure and throughput, then transmitted this data on a satellite network back to UCI’s data center. If pressure or throughput dipped, their asset management system would generate a work order and dispatch a technician for maintenance. Dexterra provided them with a platform for a mobile work order app. This was all pre-iPhone and the mobile device needed to be intrinsically safe. Smartphone electric discharge lighting up a natural gas pipeline is no bueno.
Later that year Intel Capital hosted several portfolio companies at Halliburton. It was pretty amazing to see all the IoT applications used in oil & gas production way back then. Now 20 years later, I can’t help but wonder how much better those IoT use cases will be with artificial intelligence.
By 2006 we were working with energy providers like PG&E and Puget Sound Energy. In these cases the energy was both electricity and natural gas, but with retail distribution to end users. Same story: mobile workforce apps to for linemen and field technicians to respond to power outages and maintenance. Same wonder: how can these processes automate even more with machine learning and artificial intelligence?
Here’s one last personal anecdote. My good friend Sterling Lapinski got into energy trading after we graduated from Wharton. A start in fixed-income derivatives led to energy derivatives which in turn led to a realization that energy traders lacked good data analytics comparable to what he had on his Bloomberg terminal for fixed-income. Sterling started Genscape to solve that problem. He hired my other good friend Scott Brion to install IoT sensors on the power grid, transmitted the data over 2G cellular networks, aggregated the data, and sold it to traders as a service. Genscape now operates the world’s largest private network of in-field monitors and distributes industry-leading alternative energy data, delivering market intelligence across the commodity and energy resources spectrum, including power, oil, natural gas, natural gas liquids, agriculture, biofuels, and maritime freight. He then ran the same play again with ClipperData, this time focusing on oil & gas import/export data.
There is a common theme in all four of these examples. Large and unique data sets generated by IoT devices are incredibly valuable for generative AI use cases. Gas compressors, gas meters, electric meters, and telemetry on the grid all create lots of data. In many cases, this data is time sensitive and AI inference needs low latency for real-time decision making. As I write this, my wife is waiting for me to watch another episode of Landman with her. Pretty sure there’s going to be another a fireball at a rig. Seems like Billy Bob Thornton and his crew could use a lot more IoT, predictive AI solutions, and regular maintenance!
The combination of Hedgehog Open Network Fabric and Hedgehog Transit Gateway helps energy companies improve the performance, scalability, security, and flexibility of their AI strategies. These technologies ensure that data flows efficiently between edge devices, on-premise systems, and cloud infrastructure, enabling real-time AI-driven decision-making for applications like grid optimization, predictive maintenance, energy forecasting, and renewable energy integration. By providing a robust network infrastructure, they support energy companies in leveraging the full potential of AI to optimize operations and drive innovation.
This term “the edge” has driven me nuts for a long time. It’s a vague, overloaded term. So before I start using it with a lot of other jargon, let’s define it. Edge refers to the physical location or devices where data is generated, collected, and initially processed. In energy companies and the utilities industry, the edge is typically situated close to the operations or infrastructure—such as power plants, substations, transmission lines, renewable energy generation facilities, meters, or industrial equipment.
The edge could include:
Smart meters: Devices at customer premises that measure energy consumption.
Grid sensors: Devices monitoring voltage, current, and frequency across the electrical grid.
Wind turbines, solar panels, and other renewable energy systems: These generate continuous streams of data regarding performance, efficiency, and environmental conditions.
Oil and gas exploration and production equipment: Sensors and monitoring systems on drilling rigs, pipelines, and refineries.
Pumping stations: Located in transmission lines to monitor the movement of oil, gas, or electricity.
In these cases, the edge is the point where the raw data is gathered before any significant processing is done, typically within a device or local gateway. Hedgehog enables this with edge computing networked with cloud computing through our gateway.
Edge computing refers to the practice of processing data locally (at or near the source) instead of sending all the raw data to centralized cloud servers or data centers. Edge data can feed AI inference running on a small rack of servers all networked with the Hedgehog Open Network Fabric and connected to the cloud with the Hedgehog Gateway. The combination of edge computing and public cloud is often called “hybrid cloud.” Processing data close to where the data is generated—at the edge of the network - enables faster decision-making, reduces data transmission delays, and minimizes network congestion.
In the energy industry, edge computing is a key enabler for optimizing operations in real-time, as the energy sector depends heavily on continuous monitoring, control, and dynamic adjustments. The main goal of edge computing is to perform initial data processing, filtering, and analysis locally, which can improve efficiency, reduce response times, and enhance system reliability.
Real-Time Data Processing: Hedgehog edge computing allows for instantaneous analysis and decision-making. For example, if a sensor detects a voltage fluctuation or a pipeline leak, edge devices can immediately process this data and respond—such as by shutting down equipment, triggering alarms, or rerouting power—without waiting for cloud-based analysis. This can reduce downtime and prevent costly disruptions.
Reduced Latency: Data generated by sensors and devices in energy systems (such as smart grids, wind turbines, or oil rigs) needs to be acted on quickly. Hedgehog edge computing processes this data near the source, reducing latency (the time it takes for data to travel from the source to a central server) and enabling faster responses.
Efficient Bandwidth Usage: Sending vast amounts of data from energy assets (e.g., turbines, smart meters) to the cloud or a centralized data center can overwhelm communication networks, especially when the data is continuous and high-volume. Hedgehog edge computing filters and processes data locally, sending only necessary or aggregated information to centralized systems. This reduces the need for large amounts of bandwidth and allows for more efficient use of network resources.
Improved Reliability and Fault Tolerance: Since Hedgehog edge computing processes data locally, energy systems can continue functioning even if there are communication network failures. For example, if a connection to a central control room is lost, edge devices can still make critical decisions based on local data, ensuring that essential functions (such as grid stabilization) can continue without interruption.
Enhanced Security: By processing sensitive data closer to the source, Hedgehog edge computing reduces the need to transmit sensitive operational data across potentially insecure networks. This can mitigate risks related to cybersecurity, such as data interception or manipulation during transmission. Edge devices can also incorporate security protocols to ensure data integrity before it reaches central systems.
Cost Reduction: Instead of relying on cloud or centralized systems to process and store all the data, Hedgehog edge computing reduces the need for expensive data transmission and storage by handling the bulk of data processing locally. It also minimizes the costs associated with cloud services, particularly for high-volume, real-time operations.
Integrating artificial intelligence (AI) inference with IoT and Hedgehog edge computing significantly enhances the effectiveness of utility companies and energy producers in managing their supply chains, sustainability and operational efficiency. Gen AI can process and analyze large volumes of data generated by IoT devices, making it possible to extract meaningful insights, predict outcomes, and make intelligent decisions in real-time, often at the edge. Here are several digital transformation use cases where power generation and utility companies work better with AI inference powered by Hedgehog edge compute and hybrid cloud:
AI algorithms, such as machine learning models, can be deployed at the edge to perform real-time data analysis. For example, AI can analyze the data from sensors to predict equipment failures, identify abnormal conditions, or optimize energy distribution instantly. This enables automated decision-making without waiting for data to be sent to a central server.
Here’s an example. In a wind turbine, IoT sensors capture data on temperature, vibration, and pressure. Generative AI models at the edge can analyze this data in real time to detect anomalies and predict potential failures, prompting immediate maintenance or shutdowns if needed.
IoT devices constantly monitor the health and status of critical assets like turbines, transformers, and distribution lines. Edge computing processes this data locally to detect any immediate issues. AI models, especially predictive maintenance algorithms, can analyze historical and real-time sensor data to predict when equipment is likely to fail. By running inference models at the edge, AI can identify early signs of wear, prevent downtime, and optimize maintenance schedules.
Universal Compression is a great example of this use case. AI models at the edge process data from equipment sensors in real time. They can infer patterns from previous data (such as vibrations and temperature readings) to forecast when a machine part will fail, enabling a preventive maintenance action before a breakdown occurs.
These are key AI use cases for utilities companies. IoT devices such as smart meters monitor energy consumption in real time across residential, commercial, and industrial users. Edge computing helps process this data quickly, allowing utilities to track energy usage patterns and grid status without delays. AI-powered load forecasting models can then analyze past and present consumption data to predict future demand. These predictions, combined with real-time data from IoT devices, allow utilities to implement demand response strategies such as dynamically adjusting pricing, scheduling energy production, or shifting demand to off-peak times.
AI inference can help predict energy demand peaks by analyzing historical data and current consumption patterns. At the edge, AI models can issue commands to manage electricity demand—such as instructing smart thermostats or appliances to reduce energy consumption during peak hours—without needing to send data to a cloud server for analysis. This improves customer experience, energy efficiency, and sustainability. It also keeps regulators happy.
Much has been written about the need for modernization in power grids. The easiest initiatives use AI technologies in the grid to streamline use of what we already have. IoT devices on the grid (such as sensors and meters) provide real-time data on voltage, current, and operational status. Edge computing processes this data close to the source to minimize latency and ensure fast response times.
AI models can analyze the data to detect anomalies, faults, and potential disruptions in grid operations. By running AI inference at the edge, utilities can take immediate action to re-route power, isolate faults, or trigger alarms before a small issue becomes a larger problem.
AI inference can detect when an electrical component is starting to malfunction, such as a transformer showing signs of overheating or voltage instability. The AI model running at the edge can instantly trigger a circuit breaker to isolate the faulty section of the grid, preventing a larger outage.
IoT sensors monitor renewable energy sources like solar panels, wind turbines, and energy storage systems, tracking performance, weather conditions, and energy levels. Hedgehog edge computing processes this data locally for efficient energy distribution and storage.
AI models can optimize energy production and storage by predicting solar/wind generation based on weather forecasts, historical data, and current conditions. This enables energy storage optimization, ensuring that energy is stored when it's abundant and used when it's scarce.
AI models can use real-time data from weather sensors and energy storage systems to forecast the optimal times for charging and discharging batteries. By running inference at the edge, the system can immediately adjust the energy storage strategy to align with predicted demand, ensuring that renewable energy is used efficiently and reducing reliance on the grid.
IoT devices on the grid collect data on voltage, load, and equipment status. Hedgehog edge computing processes this data quickly to identify potential security breaches or malicious activities in the grid infrastructure.
AI models can be trained to recognize patterns of normal behavior and detect anomalies that might indicate security breaches, such as cyberattacks or unauthorized access to critical infrastructure. By running AI inference at the edge with Hedgehog, potential threats can be detected and mitigated in real-time.
AI models at the edge can analyze network traffic data from IoT sensors and devices for signs of hacking attempts, unusual access patterns, or compromised equipment. If the system detects an anomaly, it can trigger an immediate response, such as isolating affected devices, preventing a cyberattack from spreading.
Energy trading is a great example where data analytics can create new AI-driven business models. IoT devices gather real-time data on energy consumption, weather conditions, grid status, and energy prices. Edge computing enables real-time processing of this data for quick decision-making.
AI models can help forecast energy prices, predict demand shifts, and optimize trading strategies. By running AI inference at the edge, utilities can respond faster to market conditions, ensuring better energy price prediction and more efficient trading.
In smart cities, IoT devices monitor everything from energy use and traffic flow to water management and waste collection. Edge computing processes this data locally to enable real-time responses to city-level challenges.
AI inference helps optimize various city services by analyzing data from IoT devices in real time. For example, AI can analyze traffic data to optimize traffic lights or control the distribution of electricity and water based on demand.
AI models running at the edge can analyze traffic flow data to predict congestion and dynamically adjust traffic signals in real time, reducing emissions and energy consumption while improving traffic efficiency.
Several oil and gas companies are leveraging edge computing and private data centers to improve their operational efficiency, safety, and real-time decision-making. These technologies help manage the vast amounts of data generated by equipment, sensors, and exploration activities, enabling better management of resources, predictive maintenance, and enhanced safety protocols. Below are some oil and gas companies that use these technologies:
ExxonMobil uses edge computing for real-time data processing in its exploration, production, and refining operations. For example, in offshore oil rigs, sensors monitor temperature, pressure, and flow rates. Edge devices process this data locally to ensure that operations are optimized in real time and to detect early signs of equipment failure. In offshore oil and gas exploration, ExxonMobil uses edge devices to analyze seismic sensor data and drilling parameters to optimize operations and avoid costly downtime.
ExxonMobil also relies on private data centers for storing and analyzing sensitive operational data, such as geological data, reservoir performance, and environmental factors. These data centers help ExxonMobil comply with industry regulations and safeguard proprietary information.
BP integrates edge computing at its oil exploration sites and refineries to process real-time data from various assets, such as pumps, turbines, and drilling rigs. By performing analytics at the edge, BP can detect anomalies, optimize production, and reduce operational risks. BP uses edge computing to optimize oil rig operations by analyzing sensor data at the site itself, allowing engineers to adjust operations immediately to prevent equipment failure and improve efficiency.
BP operates private data centers for processing and storing sensitive data, including operational metrics from their upstream and downstream operations. The private infrastructure helps BP protect critical information from cyber threats and meets regulatory compliance requirements.
Shell employs edge computing in its offshore platforms, drilling sites, and refining operations to monitor equipment performance, detect early signs of failure, and optimize energy production. For example, real-time data from sensors is processed at the edge to improve predictive maintenance, optimize drilling performance, and prevent costly operational delays. In their upstream operations, Shell uses edge computing to analyze pressure, temperature, and vibration data from drilling equipment to predict failures and reduce the risk of downtime.
Shell uses private data centers to securely manage large volumes of data related to production and exploration, including seismic data, real-time monitoring of equipment, and analysis of drilling parameters. This helps Shell maintain control over its data and meet regulatory requirements.
TotalEnergies has adopted edge computing to enhance the monitoring and management of its offshore rigs and onshore refineries. Edge devices process data from IoT sensors to improve real-time decision-making, reduce operational costs, and enhance safety protocols. TotalEnergies utilizes edge computing to monitor and analyze performance data from renewable energy installations, such as offshore wind farms, as well as traditional oil and gas assets.
TotalEnergies relies on private data centers to store and analyze vast amounts of operational data, including geophysical and seismic data, and to run complex analytics and machine learning models for optimizing resource extraction and predicting maintenance needs.
Chevron uses edge computing for real-time monitoring and data analysis at its drilling sites, production facilities, and refineries. For example, edge computing helps in managing oil rigs by analyzing data on equipment health, flow rates, and environmental conditions. Chevron uses edge devices at its offshore production platforms to monitor equipment performance in real time. These systems can trigger alarms and automatically adjust settings based on sensor data, reducing downtime and improving safety.
Chevron employs private data centers for storing operational data from its oil and gas exploration, production, and transportation activities. This infrastructure supports secure processing of sensitive data and is used for advanced analytics and operational intelligence.
ConocoPhillips uses edge computing for real-time data collection and monitoring in remote oil fields and offshore platforms. Sensors and IoT devices monitor everything from drilling equipment performance to environmental conditions. Edge computing enables faster, more efficient decision-making, reducing the time between data collection and action. Edge devices at ConocoPhillips’ drilling sites process data on equipment performance and environmental conditions in real time, triggering immediate responses to prevent equipment damage or failure.
ConocoPhillips maintains private data centers to store, process, and analyze large amounts of data from its upstream and midstream operations, ensuring secure, compliant data management and running advanced AI models for optimization.
Equinor uses edge computing to monitor real-time drilling and oil extraction data, processing information locally at offshore sites to optimize operations and improve safety by predicting equipment failures. Equinor uses edge computing in its offshore oil rigs and onshore refineries for predictive maintenance, asset monitoring, and data analytics. By processing data at the edge, Equinor reduces latency and can take immediate actions to optimize production, improve safety, and minimize downtime.
Equinor has established private data centers to securely handle large volumes of operational data. These data centers manage sensitive data related to reservoir modeling, production forecasting, and geophysical analysis.
Halliburton uses edge computing to enhance drilling and exploration operations, especially in remote areas. Real-time data from drilling rigs and exploration sensors is processed locally to ensure fast decision-making, improving drilling accuracy and preventing costly delays. Halliburton’s EDGE platform integrates edge computing to monitor drilling equipment and geophysical data in real-time, enabling automated adjustments during the drilling process for efficiency and safety.
Halliburton operates private data centers to store and process sensitive exploration and drilling data. This data is used for advanced analytics, simulation, and optimization of oil extraction processes.
Weatherford leverages edge computing in its downhole tools and oilfield services to gather and process data directly at the wellhead or offshore platform. This allows for real-time adjustments to drilling operations and predictive maintenance. Weatherford’s intelligent systems leverage edge computing to collect sensor data on pressure, temperature, and other factors during drilling, processing the data locally for immediate action.
Weatherford uses private data centers to store and analyze extensive operational and performance data, providing insights that improve the efficiency and safety of oilfield operations.
Petrobras, Brazil’s largest oil company, uses edge computing for real-time monitoring of its offshore platforms and refining processes. This allows them to make real-time operational adjustments, preventing costly equipment failures and improving energy efficiency. Petrobras uses edge computing to manage real-time data on oil well conditions, adjusting drilling parameters to optimize production rates and reduce wear on equipment.
Petrobras operates private data centers to handle critical data related to its exploration and production operations, ensuring the security of sensitive data and facilitating long-term analytics.
Duke Energy utilizes edge computing to improve the management of their smart grids and optimize the distribution of electricity. Edge devices are deployed to monitor grid conditions (e.g., voltage, frequency) in real time, allowing Duke Energy to make immediate adjustments to power flow and minimize disruptions. Duke Energy’s grid modernization initiatives use edge computing for real-time grid optimization, such as detecting outages and rerouting power, while private data centers handle long-term analytics and performance forecasting.
Duke Energy also relies on private data centers to store and analyze large volumes of operational data from their smart grids and other infrastructure. This enables secure, centralized analytics for grid management, asset monitoring, and predictive maintenance.
PG&E uses edge computing to manage and monitor its distribution grid and renewable energy sources. Sensors and IoT devices installed on the grid collect data on power usage, system conditions, and environmental factors. This data is processed locally at the edge to provide real-time insights and to enable automated responses to grid events. PG&E’s smart grid infrastructure uses edge computing for outage detection and demand response, while private data centers are used for handling data related to grid reliability, renewable energy integration, and regulatory compliance.
PG&E operates private data centers to manage sensitive customer and operational data, as well as to store information on energy consumption, billing, and grid performance. These data centers also support advanced analytics for predictive maintenance and energy distribution optimization.
SCE uses edge computing to enhance the operation of its smart grid and optimize power distribution. By deploying edge devices throughout their infrastructure, SCE can process data locally, improve response times, and monitor power generation, consumption, and equipment health in real-time. SCE leverages real-time analytics at the edge to monitor electrical distribution and automatically reroute power during outages, while their private data centers support the management of grid reliability and renewable energy integration.
SCE relies on private data centers to securely manage the vast amounts of operational and customer data generated across its systems. These data centers support real-time decision-making, long-term forecasting, and data analytics for grid health and energy demand optimization.
Xcel Energy uses edge computing for grid monitoring and management to help optimize the distribution and use of energy. Edge devices process data from smart meters, transformers, and other grid assets, enabling the utility to detect faults, adjust load balancing, and respond to grid conditions in real time. Xcel Energy has integrated edge computing into its smart grid systems to monitor power quality and efficiency, while its private data centers manage grid analytics and long-term energy demand predictions.
Xcel Energy operates private data centers to handle secure customer data, monitor energy consumption, and run long-term data analytics for asset management, load forecasting, and system optimization.
National Grid employs edge computing in its electricity and gas networks to monitor real-time data from power plants, transformers, and distribution lines. Edge devices help optimize grid performance and detect faults more efficiently, ensuring that the energy supply is reliable and stable. National Grid’s smart grid infrastructure uses edge computing to perform real-time monitoring and adjust grid operations automatically, while private data centers are used for processing data related to energy consumption patterns and asset reliability. National Grid maintains private data centers to store and analyze operational data, perform predictive maintenance, and support decision-making for long-term grid infrastructure planning.
Enel’s digital energy services rely on edge computing for optimizing the operation of renewable assets and energy storage systems, while their private data centers enable data-driven decision-making for grid optimization.
Enel uses edge computing for smart grid management and renewable energy integration. Edge devices process data locally, which allows Enel to monitor solar farms, wind turbines, and electric vehicle charging stations in real-time and adjust operations based on immediate environmental conditions and energy demand.
Enel operates extensive private data centers that manage data from smart meters, grid assets, and renewable energy installations. These data centers are used for storing and analyzing operational data and running advanced analytics on energy efficiency and grid reliability.
Iberdrola’s smart grids use edge computing for local power monitoring, while their private data centers provide analytics to improve the integration of solar and wind energy into the grid. Iberdrola has integrated edge computing into its smart grid operations, particularly in renewable energy management and distribution networks. By processing data locally, Iberdrola can quickly respond to grid conditions, optimize energy distribution, and ensure energy reliability.
Iberdrola relies on private data centers for secure data management, storing information related to energy production, consumption, and grid infrastructure. The data centers also support their renewable energy and grid optimization initiatives.
PPL Electric Utilities utilizes edge devices to manage outage detection and load balancing in real time, while private data centers handle data for predictive analysis and grid health monitoring.
PPL Electric Utilities uses edge computing in its smart grid system to monitor real-time data from various sensors and smart meters. This helps PPL detect power outages, forecast demand, and optimize the performance of its grid.
PPL operates private data centers for data storage, handling customer and grid data. The private data centers support predictive maintenance, long-term load forecasting, and grid management applications.
Florida Power & Light uses edge devices to monitor grid conditions and energy consumption locally, enabling faster responses to outages and improving energy distribution efficiency.
FPL uses edge computing for smart grid management to ensure that data from their sensors, smart meters, and grid infrastructure is processed locally, enabling real-time decision-making for grid stability and operational efficiency.
FPL relies on private data centers for storing and processing operational data, managing customer billing and energy consumption, and running analytics to optimize grid performance and future energy demand forecasts.
Con Edison uses edge computing to monitor and control energy usage and grid health in real time, while their private data centers facilitate data-driven decisions for long-term energy planning and system improvements.
Con Edison uses edge computing to manage real-time monitoring of grid infrastructure, including equipment performance, electrical load, and environmental conditions. This data is processed at the edge to optimize power distribution and respond to grid events without relying on centralized cloud processing.
Con Edison operates private data centers for secure handling of operational data, energy analytics, and long-term storage. The data centers also support demand forecasting, outage management, and energy system optimization.
If you want to sustain your energy operations or build new hybrid cloud infrastructure for artificial intelligence and digital transformation projects, schedule a call with the Hedgehog team. We would love to help you reach your project goals.