How AI is building better gas stations and transforming Shell’s global energy business

Imagine a man lighting a cigarette while he’s waiting at the pump for his car to finish fueling at a Shell gas station in Singapore, unaware that with one move he could cause a fire or explosion.

An onsite video camera captures the scene, and a device inside the station running Microsoft Azure IoT Edge can now use artificial intelligence tools to pick out that behavior — out of all the cars coming and going, drivers cleaning windshields, customers buying snacks — as a potential safety risk.  It’s a first line of defense on the “intelligent edge,” where data is quickly processed close to where it’s collected, without accessing the cloud, and simple machine learning algorithms can dispense with anything that’s not of interest. They can also be trained to look for other high-risk incidents: people driving recklessly, theft, improper fueling.

Questionable frames are immediately uploaded to the Microsoft Azure cloud, which can power more sophisticated deep learning AI models. These can identify that the man is smoking and raise an instantaneous alert on an onsite dashboard, so the station manager can take action to shut down the pump before any harm is done.

This pilot project — now in action at two gas stations in Thailand and Singapore — is just one example of how Shell is integrating AI, cloud computing and Internet of Things (IoT) technology across all dimensions of its energy business. From oil and natural gas fields to the gas pump to electric charging stations, Shell has developed leading-edge technologies that are making operations safer, boosting efficiency, saving money and helping employees communicate and share solutions across the global company.

To accelerate that digital transformation on a global scale, Shell has selected C3 IoT and Microsoft Azure to power a new companywide AI platform. It plans to deploy the Shell AI Platform on a broad set of applications, from predicting when hundreds of thousands of pieces of equipment in offshore production hubs, refineries or wells need maintenance before problems arise to making sure parts and inventory can reach remote locations quickly.

“The new possibilities in working with data over the last few years are unlocking amazing opportunities in all aspects of what we do in the company,” said Yuri Sebregts, Shell’s executive vice president for technology and chief technology officer. “Most immediately, this will help us scale solutions we’ve been developing.”

Shell’s new AI platform running on Microsoft Azure will help monitor and predict when critical infrastructure needs maintenance. Photo courtesy of Shell.

“In one example, we can now forecast in many cases when a compressor is at risk of failure 24 or 48 hours in advance, which was not easy to do before despite all the instrumentation you have on these large and complex machines. We’ve proven that on several tens of compressors and to scale that to tens of thousands of machines around the world is immensely helpful,” said Sebregts.

The C3 IoT and Microsoft partnership allows companies to ingest and easily visualize real-time data from across their business operations, and it allows users to pick from a broad array of intelligent capabilities and solutions. Eventually, Shell plans to apply the platform to other machine learning, computer vision and natural language processing challenges.

It’s just one of the ways that Shell and Microsoft are collaborating to help digitally transform a company that employs 85,000 people across 70 countries, manages critical energy infrastructure across the globe and operates 44,000 retail gas station and convenience store sites worldwide.

In addition to professional data scientists, Shell also has thousands of employees who are experts in making drilling more efficient, undersea exploration or developing alternative fuels. They need a data and AI platform that is simple to use and allows them to innovate quickly, said Rohan Kumar, Microsoft’s corporate vice president for Azure Data.

“They want to be able to collect data, use this platform, have magic happen under the covers and come out with the insights they want. The investments we’re making in Azure and AI are really about putting ourselves in the shoes of that customer,” Kumar said.

‘Opportunities are endless’

At its retail sites, for instance, Shell was searching for a solution that could identify potential safety risks in the massive stream of data captured by closed-circuit video cameras — and quickly enough for employees to respond to and eliminate potential problems. Technologies offered by other cloud providers, which uploaded and processed the entire volume of video data in the cloud, took too long to be useful.

Because Microsoft’s solution uses Azure IoT Edge, Azure IoT Hub and Azure Databricks to process much of the data onsite and only uploads frames that need more advanced AI processing, nearly all that lag time is eliminated. In the pilot project, the employees are also able to provide feedback on how well the system performed, allowing the model to learn from each experience.

Shell employees are using Microsoft’s intelligent productivity tools to stay connected to peers around the globe, including those working in remote locations. Photo courtesy of Shell.

“Each of our retail locations has maybe six cameras and captures something in the region of 200 megabytes per second of data. If you try to load all that into the cloud, that quickly becomes vastly unmanageable at scale,” said Daniel Jeavons, Shell’s general manager for data science. “The intelligent edge allows us to be selective about the data we pass up to the cloud.”

Azure Databricks also allows Shell’s data scientists to use popular open-source technologies that, on their own, can be labor intensive and require a lot of support. The Databricks installation is stable and easy to use, and allows for more seamless collaboration, Jeavons said.

Shell sees vast potential for intelligent computer vision tools to automatically detect behaviors or conditions in video camera footage and alert the company to potential safety risks. It could be deployed on construction projects to flag when employees aren’t wearing proper safety equipment or to inspect equipment sitting on the seafloor thousands of feet underwater.

“Historically, we’ve had robotics with cameras that we could send down there, but we’d have to manually look through that video footage to determine if there’s a reason to intervene,” Sebregts said. “With video analytics we can completely automate that. It’s faster, cheaper, better and safer — the use cases and opportunities are endless.”

Product experts teach machines complex tasks

Shell also is collaborating with Microsoft engineers on a project that uses AI and machine learning to improve how it drills horizontal wells. The oil and gas industry has realized significant cost efficiencies, reduced its footprint and reached new oil and gas resources onshore by substituting traditional methods of drilling wells straight down for long-reach, horizontal drilling.

“Geosteering” is the challenge of pinpointing the long, horizontal well position exactly and steering it to reach the most productive rock containing oil and gas. While Shell’s remote drilling technologies have advanced geosteering, it remains complex and labor intensive, requiring 24/7 decision-making from highly skilled professionals.

Shell developed Shell Geodesic™ in house with their own AI research scientists. It’s an AI solution to improve the accuracy and consistency of a horizontal well’s directional control to reach optimal oil and gas reservoirs. Shell Geodesic™ features a drilling simulator, which offers a user-friendly interface, and a suite of machine learning and control algorithms that gives geologists and drillers a better view of the best layers of oil and gas rock.

YouTube Video

In one part of the solution, they applied a machine teaching framework developed by Bonsai, which was acquired by Microsoft last summer, that allows subject matter experts with little background in data science or AI to tell the system what it wants the intelligent agent to do and what key information it needs to know to do that job successfully.

This Microsoft team works on combining this subject matter expertise with deep reinforcement learning — a branch of AI that enables models to learn from experiences much like a person does, rather than from meticulously labeled data.

The Bonsai platform performs much of the machine learning mechanics in the background — translating instructions into algorithms, creating neural networks and teaching the model the desired behavior. Using this approach, it produced an intelligent agent that, in a proof-of-concept test, learned how to optimally steer the drill using a simplified simulated 2D virtual well environment.

“What excites us about Bonsai is that it gives us a reinforcement learning platform that allows us to scale quickly and takes away the engineering effort involved in stitching together the open-source capabilities so our data scientists can focus on what they’re best at, which is figuring out what the model needs to do,” Jeavons said. “It’s early days still, but we’re extremely excited about the potential.”

Improving employee engagement

But Shell’s digital transformation isn’t just limited to its physical wells, pipelines and plants. It’s also changing the way employees working around the globe communicate with each other.

When Shell’s internal communications team started looking for ways to boost employee engagement and empower everyone across the organization to share information, they settled on a combination of intelligent tools offered as part of Microsoft Office365: Yammer, Stream and SharePoint Online.

Leaders started using Stream, an enterprise video service, to connect with employees more authentically and personally. Now, in addition to leadership communications, employees can easily find or create videos to promote safety, share best practices or analyze a successful project. Stream features like automatic closed captioning and deep search ensure communications are accessible and help employees quickly find the most useful content.

Those videos can be easily posted on SharePoint, a collaboration repository, and Yammer, a corporate social network that allows employees to have conversations with peers across the organization and give leaders insights into what employees are experiencing. More than three-quarters of Shell employees now use Yammer, with an average of 4,000 joining each month. The discussions help unify teams that are dispersed across the globe, solve problems together and foster open communication between groups that had little contact before.

For instance, employees working the night shift on a rig off the cost of Australia might use Yammer to alert the incoming crew to any issues they’ve experienced, and they can now ask if someone working at another location around the world might have a solution.

“These tools allow people to connect with each other, to learn from each other, to see opportunities quicker and build off of each other’s skills,” Sebregts said. “I lead a global organization, and in the past someone doing my type of job might travel around the world and hold a traditional town hall everywhere and once a quarter they would send an email with some thoughts. This is a new era of communication — it’s open, instantaneous, it’s modern, it’s fast, and I love it.”

Related:

Jennifer Langston writes about Microsoft research and innovation. Follow her on Twitter.