Vertical Markets Spotlight: Speech in Manufacturing and Distribution

On the surface, manufacturing might seem like an unusual use case for speech technology. In reality, though, speech technology has already been impacting the industry in positive ways for decades and has the potential to have more dramatic impacts in the future.

There has been significant advancement in speech technology over the past few years with new innovative ways of using technology—even in industries like manufacturing and distribution. Consider how speech technology might be used in sales and service, learning and development, even during production processes themselves, for instance.

The potential of speech technology in manufacturing is exciting.

As Vikrant Tomar, founder and chief technology officer of Fluent.ai, reports in Industry Today: “Speech recognition solutions not only improve inefficiencies in factory workflows to increase production efficiency and reduce manufacturing costs but also boost job satisfaction, which improves employee retention and, in turn, reduces more cost.”

And Fluent.ai knows a little something about that. German appliance manufacturer BSH Hausgeräte recently partnered with Fluent.ai to bring speech-to-intent AI to assembly lines at its factory in Bavaria to improve worker efficiency and ergonomics.

Company officials said they chose Fluent.ai’s software for its ability to understand voice commands, even in a noisy factory environment.

For BSH, Fluent.ai’s voice recognition system lets machine operators at each workstation speak a wake word followed by a command into a headset to move products through the assembly line.

The hands-free technology has reportedly increased employee productivity by up to 100 percent.

Developers and companies are growing increasingly creative in terms of considering the myriad ways that speech technology can create greater efficiencies, drive out costs, and positively impact both safety and sales.

Vocodia’s use of AI voice technology in industry, manufacturing, and distribution helps companies “assert sales certainty to match up with supply side of distribution,” says Brian Podolak, CEO of Vocodia. Reliance on humans, he says, is far less effective and efficient.

“Vocodia’s software allows companies to address everything from simplifying ordering and processing, speeding up the entire ecosystem within and between organizations from supply and sales to security, and working with humans to operate at the speed of thought,” he says.

Chatbots are one example of how voice technology is being used in manufacturing to achieve efficiencies.

“Chatbots can play an important role in automating tedious and repetitive manufacturing processes and eliminating human error,” says Beerud Sheth, cofounder and CEO of Gupshup.

“For example, integrating chatbots within an [enterprise resource planning] system helps manufacturers improve inventory management, reduce waste, and speed up fulfillment,” Sheth says. “When integrated on the shop floor, proactive delivery notifications help keep all team members in the loop. Chatbots are being used to improve customer service and free up staff from answering common questions. We are seeing many businesses use chatbots to provide instant responses to basic queries and escalate more challenging requests to a human agent for help, allowing the latter to focus on higher-value tasks.”

Verbal workflow management is another speech tech improvement opportunity for manufacturers of all kinds. Think of it as an always-on source of information and instructions to guide any task—far more effective and lasting than traditional training.

Bill Crose, CEO of Adyton, a tech firm that offers hands-free, interactive products and systems, explains, using the role of bartender as a relatable example.

Bartenders can’t possibly memorize how to make every drink that exists, therefore their job is “unlearnable,” Crose says. So what do they do? “Bartenders commonly use their phones and YouTube, printed recipes in books, or a recipe on their POS when they need to make a cocktail they don’t know how to make.” But, he adds, because they work with their hands, “each of those aids are highly inefficient and unsanitary.”

But what if they could, instead, follow step-by-step, voice-delivered instructions? That example, Crose says, could be applied to any manual worker in any industry in the world.

“People with basic skills can perform without training when given verbal instructions, step-by-step, on demand,” he adds.

This could encompass people like nurses, assemblers, packagers, scientists of all disciplines, or engineers, Crose says.

This clearly has a significant impact on learning and development in manufacturing settings and related impacts on product quality—a minimization of defects—and safety.

“Learning requires memorization, and many things cannot be memorized,” Crose says. In the manufacturing world this can include such things as standard operating procedures (SOPs) with more steps than anyone can memorize, steps that change frequently, or steps that are performed too infrequently to remember, he says.

There’s still room for growth here, though.

James Kaplan is CEO and co founder of MeetKai, a metaverse company with conversational AI at its core. “A lot of speech technology has been used primarily for command-based interaction that is forced to follow a template,” Kaplan says. “I have not seen mass adoption of this because, frankly, command-based systems with speech fail too often since they require the user to remember the patterns.”

It’s easier, he says, “for users to carry out user interface through touch or physical interaction compared to speech if the speech is likely to fail.”

But the manufacturing environment is ripe for the application of speech technology to a wide range of operational issues with opportunities for significant efficiency improvements and related cost savings.

Consider the warehouse floor, where the market for voice-directed systems is expected to expand around 12 percent annually through 2031 as quantifiable productivity and accuracy gains are recognized, Persistence Market Research concluded recently.

Voice-directed warehouse applications are improving distribution efficiency across industry verticals, from food and beverages and grocery distribution to apparel and pharmaceutical supply, the firm found.

The research firm expects the retail, e-commerce, and consumer packaged goods sectors to offer high opportunity during the forecast period, as demand for automated mobile tasks and improved workflows is increasing significantly. Overall, these sectors are projected to account for around half of the total market share.

As the world is becoming more digital and connected, demand for automated solutions will keep increasing. Voice-directed warehousing solutions can typically eliminate nearly 80 percent of the errors caused by picking the wrong product or an incorrect quantity by humans, according to the research firm.

Speech technology, of course, also has the potential to address the staffing challenges that manufacturers face with the Great Resignation.

“The future of voice-delivered instructions can be found in the U.S. Patent and Trademark Office’s productivity numbers,” Crose says. In 1979 they granted 52,000 patents—in 2019 they granted 319,000, he says. At the current rate, he says, 40 million patents will be granted over the next 40 years.

“With each patent will come at least one job and each of those jobs will be for manual work—work that hasn’t matured to the point of automation,” Crose says. “Every one of those jobs will begin with a manual worker and no SOP. Without an SOP, there is no process to train, learn, or automate—they must all be incrementally developed and optimized. Once optimized, they can be automated.”

Voice-delivered work instructions, Crose says further, can be edited and deployed globally in seconds, unlike videos, printed instructions, or illustrations, which can take weeks or even months. Voice offers a real-time, readily accessible source of information about whatever workers might need to know in the moment.

Krishna Rungta is founder of Guru99, a media outlet that focuses on providing free training tutorials and videos for IT courses. “The benefits of speech technology are clear, and it is likely that we will see even more adoption of this technology in the future,” Rungta says. “However, there are still some challenges that need to be addressed. For instance, background noise can interfere with the accuracy of speech recognition, and some accents are still not well represented by current speech recognition systems.”

XR, or extended reality, is poised to power the adoption of speech tech in manufacturing and other industries in the future.

“The cutting edge and where I see the future of speech tech in manufacturing is at the intersection of XR and conversational language,” Kaplan says. There is a lot of work in progress, he says, including at MeetKai, “to use speech and AR to overlay information for workers in a manufacturing environment. Workers can then use speech and gestures to take action or update the type of information being shown.”

XR, Kaplan says, effectively gives employees “superpowers.” The main barrier right now, though, he says, “is to get costs down and make sure that the systems are intuitive enough that they never reduce efficiency.” 

Linda Pophal is a freelance business journalist and content marketer who writes for various business and trade publications. Pophal does content marketing for Fortune 500 companies, small businesses, and individuals on a wide range of subjects, from human resource management and employee relations to marketing, technology, healthcare industry trends, and more.