Sunday 3 December 2023

IBM rolls out McDonald’s drive-thru chatbots with NLP technology under the hood

By Radithebe Rammutle   

On May 30, 2022, IBM announced that it will be rolling out more McDonald’s drive-thru chatbots across its 14 000 United States (US) restaurants. Officially known as Automated Order Taking (AOT) units, the latest IBM-powered chatbots are an improvement on the piloted version developed by Silicon Valley voice technology start-up Apprente. Apprente piloted this technology in ten McDonald’s drive-thrus in Chicago and the results were promising. The units installed were accurate by 85% and only about 20% of orders required employee intervention. 

McDonald’s has been the market leader in its industry and has developed solutions that have now become industry standards. However, in recent years its leadership has been challenged and its growth has become sluggish. Hence for the past three years, the US-based fast-food chain business has been taking bold steps to invest and reorganise its technology business. These moves are part of the erstwhile “Velocity Growth Plan” and current “Accelerating the Arches” strategic growth plan that is leaning toward technology to reinvigorate the Golden Arch’s growth. 

In 2019, under the Velocity Growth plan, McDonald’s leadership launched McD Tech Labs, a technology lab based in Silicon Valley. They simultaneously acquired Apprente for an undisclosed amount of money to lead the lab. On the occasion of announcing the deal, former McDonald’s CEO Steve Eastbrook said, “Building our technology infrastructure and digital capabilities are fundamental to our Velocity Growth Plan and enable us to meet rising expectations from our customers, while making it simpler and even more enjoyable to crew members to serve guests.”  

Despite Apprente’s progress and impressive performance, McDonald’s upped the ante and announced in the third quarter of 2021 that it was selling its McD Tech labs for yet another undisclosed amount to IBM. McDonald’s regarded IBM as more capable to scale its AOT units and to innovate with additional languages, dialects, and menus.  During the 2021 third-quarter earnings call, McDonald’s CEO, Chris Kempczinski said there was a lot of work to go into building the AOT technology. “That work is beyond the scale of our competencies if you will. And so, I think, in this case, IBM is a natural partner for us” said Kempczinski. The McDonald’s chatbots are powered by IBM’s Natural Language Processing (NLP) technologies, used to build machines that understand and respond to text and voice data and respond with text or speech of their own in much the same way humans do.  

“We believed, (and we built a thesis around), we could use our natural language processing technology, which is very good to augment the McDonald’s technology…” said Rob Thomas, IBM’s senior vice-president of global markets at the JP Morgan’s 50th Annual Global Technology, Media and Communications Conference, held on May 23-25, 2022.

IBM’s conversational technologies or NLP-driven tools and services are part of the Watson AI portfolio designed to make it easy for its customers to use data from diverse sources and trust the recommendations and predictions from AI models. Within the portfolio, there is AI for customer service, AI for business automation, NLP, and Explainable AI. 

The McDonald’s chatbots could be deployed using one or a combination of NLP products. However, the primary product in the case of chatbots is IBM Watson Assistant. This technology is a voice-based solution used to build, train and deploy chatbots over the Cloud. It is built on deep learning, machine learning, and NLP models to understand questions, find, or search for the best answers, and complete the user’s intended actions. 

IBM has three other additional NLP-driven tools and services that serve different business needs. The Watson Discovery is a tool used to extract meaningful business insights from documents, webpages, and big data. This technology is appropriate in contexts where enterprises are seeking to eliminate data silos and retrieve information through search and text-analytics programs. 

The Watson Natural Language Understanding can be used to analyse unstructured data formats such as HTML, webpages, social media, and others. This technology is often appropriate for Public Relations pros, and product developers who can no longer afford to rely on human screening to make sense of what has been said about their products, brands, and companies. Finally, IBM Watson Annotator for Clinical Data was built specifically for healthcare and life sciences domains. It extracts key clinical concepts from textual data, like conditions, medications, allergies, and procedures. So, don’t be surprised the next time you visit McDonald’s drive-thru, your order might be taken by a machine. Always remember that it is IBM under the hood when you drive through.

Older Posts

Democratisation of global payments and settlement system

The exclusion of Russia from the global financial system shows why alternatives to SWIFT such as Afreximbank’s proposed PAPSS could be necessary in the future to ensure Africa’s self-reliance to move and receive funds abroad.

If you can keep up, Cloud is the silver lining

If you don't believe Cloud is one of the catalysts to business evolution in our generation, then you haven't been paying close attention.

Innovation of hackers on the rise

On 14 August 2003, residents of two major cities in USA and Canada - New York, and Toronto - could never have predicted the blackouts that were about to befell them.

Enterprises UP completes a study that identifies emerging 4IR technologies and future skills needs

For the past few years, the term Fourth Industrial Revolution, often abbreviated as 4IR, has received splendid attention from the research and business community.

Outsmarted: Tech powered retail investors defeats Wallstreet in the battle over GameStop stock

Wall street hedge funds, Melvin Capital, Citron and others, rushed to close their GameStop short-positions as the share-price of the gaming retailer rose to a peak of $347,51 in January 2021 from a low of $12 in December 2020.

Sipho Mogapi Helps Businesses To Launch Market-Fit Digital Products

Without the help of luck or a silver spoon, it took Sipho Mogapi more than 10 years to find his niche in bots.


Previous Editions

subscribe to our magazine