Shell taps into AI to streamline operations and refine customer-centricity
The potential benefits of artificial intelligence (AI) are on the agenda of almost every management board. Questions are being asked about the ability of AI to automate processes, bring robots into the workforce, enhance and lower the costs of customer services and more. However, expectations for AI often run ahead of reality, and it is falling to CIOs to provide grounded guidance.
As Craig Walker, CIO of Royal Dutch Shell’s Downstream division, explains: “The problem with AI is that people have a certain vision of it. The AI that they think of most commonly involves human-like robots. But, of course, there are different layers to AI.”
Refining the benefits
That subtlety and breadth of AI’s potential is certainly reflected at Shell, Europe’s largest company. Within its Downstream division, the technology’s ability to improve efficiency and performance is already evident. “Being able to make more intelligent decisions based on learnings that we can give a system can have a significant material impact on profitability,” says Walker.
As he outlines, AI is being used in the analysis of the huge volumes of information created in the day-to-day running of its plants. Traditionally plant managers have relied on sales managers to inform them of likely future sell volumes. But Shell has found that a more profitable approach is to use AI to anticipate demand within the market and track shortages of supply that will allow it to maximize the value generated by a barrel of oil.
As well as using an intelligent system to understand market patterns, Shell Downstream is using AI to analyze the different crude oils in its plants to ensure the different mixes and blends don’t result in any problems during the refining process. That helps Downstream managers make better decisions upfront with an eye to the bottom line, he says.
That Smart Manufacturing system advises engineers each morning about any trend it has detected during the previous day, he says. “As well as highlighting any issues, the system provides options and recommendations on the best temperatures, pressures and running speeds that can help get an extra 10 tonnes or even 100 tonnes of [refined material] out each day — and that can be worth a lot of money,” says Walker.
Another of the AI technology platforms in action at Shell Downstream is Salesforce.com’s Einstein. It is able to make predictions and recommendations by digesting operational data as it passes through the systems, for example, prioritizing the order in which maintenance tasks should be tackled. “It’s helping us make faster and better decisions,” says Walker.
From back to front
Such applications may be driving greater profitability at the back end of Shell Downstream’s operations but with crude oil prices still sitting at around half their 2014 rates, Shell is also turning to AI to move the focus from volume to customer-centric commerce. By applying AI at the front-end of operations — the Downstream division supplies fuel and lubricants to businesses, aviation and shipping, as well as operating 40,000 Shell-branded service stations — the company can offer more personalized rather than generic products and services, right down to the individual.
For Walker that approach is designed to build customer loyalty, with AI used to develop new levels of customer knowledge and insight that will enhance services. An example is the Shell Connected Car Open API. Available on mobile devices and in-car systems, it can analyze user behavior and offers tailored convenience and loyalty rewards. “We can predict what a customer might want, see when they arrive at one of our retail sites, and hit their phone with an offer: ‘Great to see you back. Specially for you today, 50% off Diet Coke or free when you buy X.’ It’s about trying to upsell in the moment.” It’s also about a shift of profitability. In 2016, Downstream’s earnings totalled $7.2 billion compared to a loss of $2.7 billion in Upstream where the focus is on exploration and production.
Shell is also using similar decision-making analytics for business customers. A salesperson can access and share insight into the performance of an oil or lubricants within a customer’s machine. “We want that customer [such as an airline or shipping operator] to be happy. So we might actually suggest they buy less oil or a different oil because the analysis shows their machines are not running as hard on the product as previously expected. We’re trying to build a win-win relationship with the customer, and the analytics can help provide that added value.”
This, for Walker, is the key to future success for his team, and in turn the company. He believes the quality of the IT-enabled interactions with customers reflects the brand as a whole. “The important thing is about how IT is delivered in front of the customer. We have a very strong brand and very strong loyalty. So our IT has to look just as good. If it doesn’t, or if our people are not communicating in the right way, why should customers believe the product’s good?”
• Images: Aviation fuel - Shell; Forecourt-Getty