Deep Tech Company Using AI To Power Energy Industry’s Digital Transformation

Glenn/ Blog/ AI

July 29, 2021

< back to all blogs

Donovan Nielsen, President of Scovan, speaks alongside CEO of Drishya on the importance of AI in the Oil and Gas/ Energy Industry.

Elsie Ross
Associate Editor of the Daily Oil Bulletin

July 27, 2021

As CEO of deep tech company Drishya AI Labs Inc., Amardeep Sibia has found himself living in what he describes as two different ecosystems — Calgary’s “precise by-the-book engineering culture” and the “fast frugal software ecosystem” of Bangalore, India. “I find it fascinating how we bridge between these two ecosystems or cultures of technology,” he said in a recent interview.

The company was founded in February 2020, shortly before COVID-19 turned the world upside down. Drishya sees its role as using artificial intelligence to power the digital transformation of the energy industry in partnership with local engineering companies. “We think, behave and operate like a deep tech company which is delivering a cleantech impact,” said Sibia. “We ardently believe we are making energy intelligent and clean by adding brains to data.”

The energy efficiency gains from these transformation efforts not only reduce greenhouse gas emissions but also save money for operators. Because of the stacking of incremental energy efficiency gains, the use of AI on its own can help reduce GHG emissions by between five and 10 per cent, said Sibia. “And that’s massive without the need for capital investments.”

Drishya (vision in Sanskrit) has two products targeting the oil and gas industry — Artisan and Brains. Artisan enables what the company calls engineering digitalization, speeding up a number of engineering processes, while Brains extracts operational intelligence from large troves of collected historian datasets and enables intelligent automation.

The company also uses what it calls artificial intelligence / machine learning (AI/ML) enablement. For example, one of Drishya’s clients has a new generation steam generator that enables the operation of smaller SAGD plants. By the time this equipment goes to the field, it has an AI/ML layer on top of it. “Think of it as an embedded engineer,” said Sibia. “It is already prepared to do self monitoring, self diagnostics, self healing and self optimization from day one.” That’s important because a plant in a remote area with that capability will operate better and be more energy efficient, he said. “You don’t have to fly in people, equipment and diagnostics. AI has all of these invisible layers that may make things that much more sustainable.”

Artificial intelligence also can learn about things fast — it can teach itself. As companies move through the energy transition, they have to manage everything faster but the data isn’t there — they haven’t even had the time to learn it, said Sibia. “In today’s day and age, I look at AI as this ability of systems to understand context, being able to look at things and react like a human.”

Artisan has been taught to look at engineering drawings the way an engineer looks at them, to understand them and then to speed up a number of engineering processes by integrating data from instruments and the use of predictive algorithms. For example, it will take about 10 seconds to understand the entire diagram and in a couple of minutes it will take in all the drawings for a plant and assemble it with the knowledge of how it is set up.

“It starts creating that single source of truth and it enables you to do a number of engineering tasks very quickly,” said Sibia. For example, a fugitive emissions study that might take 50 to 60 hours for a small building can be done automatically in under a minute.

“Once you know that you have got a single source of truth, you are able to then automatically visualize the plant which is digitalized now going forward,” he said. “We are now training the system, so it’s learning, beginning to understand which are more efficient designs, which are the designs where you might have more failures or more leaks or more emissions.” The system can also identify a more cost effective design or whether the weight in a particular design can be reduced to bring down the shipping cost.

Brains and data

On the Brains side, the AI/ML enablement sits on top of the historian data. And once it takes in data and related information, “we add Brains to the data making it more intelligent.”

In a SAGD operation, for example, Drishya deployed an intelligent soft sensor for monitoring a number of aspects of the evaporator operations such as chemical readings. The inferential sensor can provide hourly, rather than daily readings, enabling faster decision-making. The ultimate objective is to optimize operations, planning to reduce downtime or the frequency of downtimes for overall capacity increases.

“The beauty of this is that we were able to deploy these without any outage and nobody going onto the site and no capital cost,” said Sibia. Historic data is dumped from the company’s archive, not from the live system, and once it’s pulled into the cloud, its modelled offline, he noted.

Added value for clients

Donovan Nielsen, president of Scovan Engineering Inc., one of Drishya’s partners, said partnering with Drishya makes sense not only for the direct service Scovan offers but because it allows it to offer more to its clients over the course of the life of their projects. “Automated processes make it faster, cheaper and better, and hopefully bring more quality on the engineering side to our clients.” In addition, with the creation of a digital twin, Scovan’s clients can take that digital platform and continue to leverage it throughout the life of their projects to optimize and improve their operations, including reducing emissions.

Nielsen said he sees great potential in the Artisan program on which his company is working with Drishya. “If we can reduce the amount of repeated tasks that our engineers do and use Artisan to start creating some engineering deliverables, that just means we can start elevating our engineers to solve more complex problems and figure out how to add more value to our clients, instead of just creating the basic engineering deliverables.”

While automated drawings are already in use through other processes, Artisan is a lot more flexible because it enables the scanning of items, creating a structured database from which it automatically creates all the other engineering deliverables, he said. “It fits the engineering workflows better. It fits the way projects are developed better and just provides more flexibility and helps us do it faster.” The company is starting to use Artisan on a trial basis and “it looks promising,” said Nielsen.

On the Brains side, Scovan has done one project with a client. “They found it was a great opportunity just because they don’t even know what data they really collect and how to structure it and organize it,” he said. “Beyond just the opportunity to optimize things and make your operations better, it actually starts to let them use their data for simple things like planning.”

Brains ties in not only what companies get from their historian such as plant operating data, but it starts to tie in other data including financial and maintenance, said Nielsen. “I think it’s going to tap all this potential to start using all this incredible data that we collected to do something with it.”

And while the gains may only be incremental, “I think what we’re chasing is incremental gains,” he said. “And [when] we start to stack all those up on top of each other, the results are impressive.”

Originally published in Daily Oil Bulletin:

www.dailyoilbulletin.com/article/2021/7/26/deep-tech-company-using-ai-to-power-energy-industr

Move Forward