< News

Smart Energy: A Blueprint for AI, IoT And 5G Convergence

Read the full Forbes article here.
Author: 

 

For scale, consider the Statue of Liberty, standing 305 feet tall. At 466 feet, the average wind turbine in the U.S. dwarfs Lady Liberty by more than half. And when GE’s next-generation monster wind turbine, the Haliade-X, hits the market in 2021, it will nearly double that size to 877 feet, just shy of the Eiffel Tower. A single Haliade-X rotor blade will stretch 315 feet, longer than a football field.

As a general rule of thumb, when it comes to energy and energy exploration, bigger is better: the larger the machinery, the deeper the dig, the greater the production yield. But this massive scale can pose major challenges. By necessity, assets like oil rigs, wind farms and mines are often located in remote and harsh environments, posing safety risks to human workers during construction, inspections and repairs. Equipment laden with sensors can collect petabytes of data, but without reliable high-speed wireless infrastructure, transmitting it can be slow and unwieldy, straining the system’s bandwidth.

That’s all changing—and fast, thanks to rapidly evolving and emerging 5G, AI and IoT technology. Collectively, they’re transforming the energy sector in fundamental ways, by enabling energy and mineral harvesting optimization, predictive and automated maintenance, high-volume and low-latency data delivery, and smarter power grid management for better allocation of energy resources to countries, cities, manufacturers and consumers.

“Together these technologies are improving efficiency, driving down costs and allowing companies in those spaces to make better use of the available assets,” says Paul Miller, senior analyst at Forrester Research. “The biggest impact is around asset visibility and asset management. What is my equipment doing? Is that wind turbine turning and how much is it producing?”

But the benefits of today’s tech advances go beyond ensuring that machinery will perform at peak capacity. Miller sees IoT and AI’s impact transforming the very concept and function of a city’s power grid, which has remained structurally unchanged since the 1930s.

“The energy grid for a city is a mix of different sources: nuclear, gas, wind and solar,” he says. “You’re going to want to use renewable energy as much as possible, but you have to make a guess for how much energy is going to be needed by the city. Essentially you want as much data as possible to make those guesses as data-driven as possible.”

“The ultimate impact will be for every city and every country optimizing their use of power so that you need to produce less,” Miller says.

Viewed up close, current tech advances in the energy and mining sectors look like a patchwork quilt of isolated improvements, but the big picture shows something more sweeping. Here are four key arenas where AI and IoT are changing the game in the energy and energy exploration sectors.

Yield Optimization


By 2020, the industrial IoT is expected to comprise more than a trillion sensors, each collecting and sharing data in real time. This mountain of data, when processed and analyzed by advanced machine learning software, will let energy companies monitor and regulate production to cut costs and maximize output—down to the minute.

A McKinsey study projects that AI innovations could save oil and gas companies as much as $50 billion in production costs annually. Among the companies innovating in the synchronization of AI, IoT and oil and gas hardware is Calgary-based Ambyint, whose intelligent “adaptive controller” platform samples data from thousands of vertical and horizontal oils wells every five milliseconds to recommend optimization strategies. San Francisco–based Tachyus also integrates real-time equipment data with seismic activityto regulate maximum oil flow through pipelines.

Predictive AI is even helping improve how oil and gas companies locate the most resource-rich drilling grounds. Chevron is using AI to identify new well locations in California; by drilling in better locations, production has risen by 30%, the oil giant claims. Recently, BP invested $20 million in Beyond Limits, an AI startup commercializing cutting-edge tools from NASA to adapt deep-space exploration technology to deep-sea oil and gas exploration in the search for promising drilling grounds.

For the wind and solar power industries, AI is enabling greater energy yield through advanced weather forecasting and analysis. How do you maximize wind power when the wind dies down, or solar power during overcast days? By incorporating intelligent “tuning” mechanisms into the hardware that automatically adjust control settings for varied weather conditions.

GE Renewable Energy is taking a different tack to optimize wind power by creating digital wind farms. These “digital twins” are virtual models of actual wind farms that gather data from the physical turbines during operations and analyze potential settings to determine optimal efficiency. GE reports that its digital-twin technology will boost energy production by 20% annually, generating $100 million more profit over the lifespan of a typical 100-megawatt wind farm.

Predictive Maintenance And Cognitive Vision


In northeast Iowa, on a blustery day in March recently, a wind turbine’s blades churned steadily. But 400 miles away, data analytics software detected an anomaly: Unexpectedly, the turbine’s gearbox was on the verge of failure. The wind farm’s operators quickly dispatched a crew for a $5,000 repair job, averting a catastrophic breakdown costing several days of downtime and $250,000 in lost revenue.

But predictive maintenance, enabled by AI and IoT, isn’t just about preventing unforeseen equipment failure. By predicting wear and tear, it allows timely maintenance that can extend the life cycle of complex and costly machinery. More importantly, it can ensure the safety of human crews who scale massive equipment while exposed to the elements or who must attempt a dangerous rescue mission after a mine collapse.

Much of predictive maintenance technology today is enabled by sophisticated IoT sensors inside machinery to monitor temperature, moisture, output flow and seismic vibrations. Externally, AI-enhanced drones and robots are proving equally valuable in revolutionizing inspections and repairs.

Among the powerful new tools for monitoring outdoor machinery such as oil rigging and wind turbines is Aerialtronic’s “digital vision” platform, a camera-computer hybrid that can be mounted to drones or mobile robots. Its optical and thermal cameras, along with an onboard 1.5-teraflop GPU, let it detect even the tiniest of fissures that could lead to equipment failure. Another digital vision system from SkySpec lets an autonomous flying drone inspect an offshore wind turbine in less than 15 minutes. If it finds damage, its analytics can project repair costs and calculate whether they’re worthwhile, or if it’s more cost-effective to replace the equipment.

Environmental And Safety Upgrades


A recent survey asked executives from 100 of the largest mineral extraction companies in the world to name their top priority for deploying IoT in mining operations and 47% of them gave the same answer: monitoring their mines’ environmental impact. The reason? Meeting strict government regulations on environmental impact is costly, but an even greater responsibility is ensuring the health and safety of miners.

Companies like Inmarsat are working with mining companies to leverage IoT and machine learning to bolster worker safety and environmental compliance using smart sensors. Wireless sensor networks provide early detection of excessive vibrations that could lead to structural collapses, as well as the presence of dangerous flammable and combustible gases such as methane and carbon dioxide. Data collected by these sensors, as well as workers’ wearable sensors and sensor-laden flying drones used to conduct site surveillance, helps mining firms generate predictive models to minimize future dangers. All told, experts predict that smart sensors could save the mining industry $34 billion in costs by reducing health and safety incidents.

The nuclear power industry is also tapping machine learning to improve reactor safety, which could strengthen it as an alternative power source in the U.S.; currently, nuclear power plants provide 20% of all electricity generated. (Nuclear power safety is no small concern. Since the 1986 Chernobyl disaster, 56 of 99 major nuclear power accidents have occurred in the U.S.)

Engineers at Purdue University have created a deep learning neural network that can detect minute cracks within nuclear reactors by rapidly analyzing video images, which until now has been a lengthy, tedious and imprecise job for human inspectors. Rendering the inspection task even more difficult? Large sections of nuclear reactors are underwater and difficult to monitor.

Trained on a dataset of 300,000 images of crack and non-crack examples, Purdue’s AI has scored a 98.3% success rate in identifying tiny fissures in reactor walls—a significantly higher rate than that of human inspectors.

Autonomous Energy Production


In a report on the future of AI for the renewable energy industry, global risk management consultants DNV GL envision a day when wind and solar farms could spring up without any human involvement. Self-driving trucks could transport wind turbine and solar array components from the factory to the site. Another set of robots would unload and assemble them on a foundation dug in earth and filled with concrete by more robots. Finally, drones and robots would assemble the working facility.

Far-fetched? Not entirely. Autonomous mining is already underway in Boliden, Sweden’s Kankberg gold mine, and plans include its eventual operation without any human workers. In conjunction with the Swedish government, the mine’s operator has teamed up with telecom giant Ericsson, Volvo and Abb on the innovative project.

Self-driving excavators and haulers remove minerals from the 500-meter deep site. A 5G wireless network connects all machinery and sensors to ensure seamless production, transmitting data at 100 gigabits per second, nearly 100 times faster than current Wi-Fi technology.

No human workers mean no humans are at risk from mining accidents or disaster. A 24/7 production cycle optimizes value for mining companies. All told, the benefits of autonomous energy production are clear. Even if its arrival as a reality is far off, one by one the pieces are falling into place thanks to the convergence of 5G, AI and IoT. Together these advances are disrupting the energy sector at every stage, from production to refinement and consumption. In the coming decade, you can expect to see this sweeping digital transformation pay off in lower-cost, lower-risk and higher-yield businesses.

Read the full Forbes article here.
Author: