Can AI work in oil and gas without historical data?
In some countries in the Middle East, the oil and gas (O&G) sector contributes as much as 90 per cent to gross domestic product (GDP). Despite the global push towards sustainability and renewable energy, oil is still an incredibly important part of the region’s economy. Yet despite this, the level of “disruption” and engagement with startups in this vast sector has been limited when compared to sectors like retail or food and beverage (F&B).
This is mainly due to how heavily regulated the sector is, the high levels of sensitivity of industrial information coupled with increasing restrictions on data-sharing. This has all resulted in minimal room for entrepreneurs to engage in the oil and gas sector beyond developing cleantech solutions.
But increasing pressures stemming from a global push towards sustainability and energy conversion, adopting and implementing advanced, AI-driven practices have become imperative to ensure a sustained and cost-effective oil production process.
According to a survey by IBM Institute for Business Value (IBV), 82 per cent of the oil and gas executives believe innovation will be “critical” to their organisations’ success in the next three years.
Several O&G producers have recently amped up investments in their digital foundation and are further seeking to work with startups in a bid to acclimate to the ever-changing operating landscape in the industry. Among them is Saudi Arabia-based oil giant Aramco, which through its entrepreneurship arm, Wa’ed, has invested in startups like Medra Arabia, a nitrogen gas manufacturing company and Pure Polymer Factory which produces different coloured plastics. Abu Dhabi’s Masdar City partnered with BP in 2018 to launch the region’s first cleantech accelerator The Catalyst, whose graduates include Volts, a smart battery management system, and Solva, the world’s first electric last-mile delivery platform.
One technology that O&G players are increasingly employing is artificial intelligence. In an attempt to address the prevailing lack of data and enable an enhanced adoption of AI, UAE-based nybl tends to play AI differently. Instead of relying on historical data, nybl looks to turn underutilised data and experiences into actionable insights.
“We wanted to first address historical data because this is the biggest challenge, because historical data is biased, challenging to obtain, controlled by big companies or it doesn't exist," says Noor Alnahhas, founder of nybl.
Prior to founding nybl, Alnahhas along with his founding team, worked in the oil and gas field and so when it came to building up a product for his startup, his focus was directed at the O&G sector.
"We first do this in an industry where we know everything about oil and gas, then we'll address the platform, and then we'll address enabling anybody to do this, and so this is how we broke our goal," he adds.
In O&G, nybl helps companies predict imminent non-trading failure in oil pumps, which in turn helps them maximise the productivity and efficiency of these pumps, while ultimately reducing operating costs.
“These pumps, on average, produce 2500 barrels of oil a day, in some parts of the region, some of them produce 5000-10,000 barrels a day. This half-a-million dollar worth of piece of equipment can cause a $100 million loss when it fails unexpectedly,” says Alnahhas. “What we do is we tell them when that pump is going to fail so that they can reduce this downtime and plan ahead.”
nybl first conducted its pilot test with Abu Dhabi National Oil Comoany (ADNOC). During a six-month pilot test, Alnahhas said that they were able to save the company close to $10 million in operational costs.
"In the first two predictions, they (ADNOC) waited, and they found out they were true. By the third prediction, they were taking action on what we did. So they were actually changing parametres in wells, removing wells that were about to fail. It was an incredible handler,” says Alnahhas.
At the beginning of nybl's journey, it was a challenging feat for Alnahhas to prove his product's viability and its ability to outperform traditional AI used by global O&G producers.
“We started the proof of concept in August of last year, we went up against IBM Watson, MIT, Schneider, and Honeywell, all the big names in the tech industry in general AI, and we outperformed every single one of them,” says Alnahhas.
nybl now counts 12 O&G companies as its clients including Schlumberger, ADNOC, Baker Hughes, Kuwait Oil Company (KOC) and Borets International, and is looking to acquire more in Asia and Central and South America.
In addition to the O&G sector, nybl currently has five products that serve multiple industries, such as supply chain and healthcare, and plans to further its AI application in insurance, physical security and agriculture.
nybl also has plans to launch a platform for consumer products, enabling F&B manufacturers to access insights regarding consumer behaviour and their changing consumption patterns.
“Our big goal is to democratise the deployment of artificial intelligence solutions,” he says. “We are looking to help anybody figure out AI at any time. This is where we will provide consumer behaviour information to sellers, retailers and to large manufacturers.”
nybl is currently in the process of closing a financing round led by a cluster of international investors. Highlighting a severe lack of investor interest for homegrown deeptech startups like his, Alnahhas recounts a difficult experience with regional investors. As a result, nybl had to seek funding from investors outside the region.
"It's been such a battle talking to investors in this region. We stopped talking to any investors from the region because it didn't make any sense to us anymore," he says.
Six months of speaking with investors from the region resulted in no investment, yet the company’s last round was oversubscribed by a group of global investors.
Out of the $986 million invested in startups in the first half of this year in Mena, just $9.7 million went to startups in the AI and deeptech space according to Wamda Research Lab.
Typically, deeptech businesses require constant investment for their roll-out and commercialisation and this lack of risk capital directed at deeptech represents the biggest headwind to nybl's long-term suitability.
“We need to grow and hire people, we need to onboard clients and keep going and that takes capital,” says Alnahhas.