People have become picky eaters. Our ancestors ate whatever they could forage, but modern day Homo sapiens expect gourmet meals at street food prices on demand. To meet fickle consumer tastes, food and beverage (F&B) companies are looking to artificial intelligence to help them scale new products and stay profitable. Whether they are hacking logistics, human resources, compliance, or customer experience, these smart brands recognize the ways AI can impact how fast-moving consumer goods (FMCG) are produced, packaged, stored, distributed, marketed, and consumed. Artificial intelligence and machine learning are fundamentally changing the consumer packaged goods (CPG) and food and beverage industries.
Aside from the challenge of mounting consumer expectations, established food and beverage companies are also facing a shift in customer trends away from global conglomerates toward local, artisanal providers. Consumers are demonstrating not only a willingness to shovel out more money for a “handcrafted” experience, they’re also getting caught up in the DIY preparation trend of home cooking and craft brewing.
“CPG, in general, is facing this perfect storm, where activist investors are expecting a lot in margin while consumers expect more high-quality tailored products … along with better service,” explains Ben Stiller in an interview with TOPBOTS. (Stiller heads digital transformation and analytics for Deloitte’s Consumer Products Business.) No wonder many players in the CPG (or FMCG) space are going beyond automation to the more esoteric fields of big data, machine learning, and other aspects of artificial intelligence.
A taste for trouble
Consumers judge food based on its impact on their palate and their wallet, but successful food brands with staying power require more than just a killer recipe. Any of the following challenges regularly plague CPG companies trying to speed up and maintain innovation:
- Product design and specifications (or the recipe)
- Raw materials (or the ingredients) to create the product
- Equipment, tools, and machinery to scale production
- Venue (processing plant, factory floor, etc) where a company assembles/processes goods
- Safety and quality control implementation
- Compliance with government/international regulatory standards (health, environmental, safety, financial, zoning, etc.)
- Product packaging and tracking system
- Inventory management for storage and distribution
- Logistics and transport for distribution
- Marketing and public relations
- Long-term engagement with partners and intermediaries for sale
- Back office operations
- Sales and order tracking that follows the brand’s supply chain, manufacturing, and logistics processes
This is a long list of problems, isn’t it? In addition to minding all the possible points of failure mentioned above, food and beverage companies need to mitigate significant risks like contamination and spoilage, even when the products in question have been passed along to retailers and are no longer within their control.
Can AI be the magic elixir?
Shampoo, soda, and mayonnaise may be everyday products, but the infrastructure behind the production and consumption of CPG products is much more complicated than you may imagine.
“Once the ingredients and materials get into the building or assembly line to build the product, that’s where the challenge begins,” reveals Leading2Lean CEO Keith Barr. “Machines were designed back in the day to run a certain way. If anything doesn’t meet the exact standard to run that way (e.g., materials don’t show up in time or are out of spec) they just won’t run. Then when it stops, you have to manually stop and fix it.” Another challenge is that older factories lack sensors and tracking equipment, so these abnormalities aren’t logged and therefore continue to plague the food production process.
A developer and provider of streamlined manufacturing software and cloud-based solutions, Leading2Lean helps businesses achieve sustainable process improvements through data analytics. Using data analytics to detect and eliminate inefficiencies, the company helped Ohio-based specialty food maker Lakeview Farms achieve significant reductions in line downtime (34 percent), equipment repair costs (15 percent), and worker overtime ratio (17 percent).
In the CPG space, pressure to seek out providers of automation and AI-driven solutions, like Barr’s company, hinges on several factors:
- There are more marketing and distribution channels to engage.
- Competition has gone from brisk to brutal.
- Unified, synchronized data across all departments reduces errors, downtime, and costs.
- Visibility across all stages of the business process serves as a key competitive advantage.
- Real-time data on customer behavior and market trends helps future-proof businesses.
Dr. Tom Bradicich, vice president and general manager for Servers and IoT Systems at Hewlett Packard Enterprise, puts it in another way: “Customers can’t stop their businesses, so they are challenged with how to keep it going while improving operations all at the same time.”
Currently working with a major F&B CPG company to integrate new technologies in production, Bradicich believes automation, edge computing, and artificial intelligence are set to dramatically reduce human errors, hike quality, and increase sales. His team is currently rolling out a new product class called Converged Edged Systems that aims to establish more reliable production environments that cut costs and require less energy and space.
A buffet of automation options
The enterprise has used AI to tackle challenges ranging from gaming and dating to banking and health care. Despite the wide range of applications, F&B companies tend to stick to specific use cases, according to Lori Mitchell-Keller, global general manager of Consumer Industries at SAP.
Describing how clients are using the capabilities of SAP’s new Leonardo Machine Learning Foundation, Mitchell-Keller cited key AI applications that positively impact the front- and back-end processes of F&B companies:
- Shelf management. F&B retailers use AI to automate inventory management. One use case is to have staff take photos of store shelves to initiate a machine learning process that automatically detects missing or misplaced items and notifies stakeholders to restock or make corrections.
- Image-based procurement. AI and image-recognition technologies can ease the procurement process and reduce the time it takes to send an order. Employees can take a photo of an item to activate an automated database search for the exact item or an equivalent product.
- Personalized customer service. Using chatbots or voice assistants powered by natural language processing and machine learning, companies can tap consumer shopping data and history to provide hyper-personalized and automated customer service experiences.
- Heightened consumer engagement. CPG players can use AI to maintain strong empathy with their audience. By closely monitoring conversations on social media, companies can use AI to analyze consumer data and identify sentiments or behavior that are crucial not only in building positive experiences but also in the development and design of new product lines.
Some CPG businesses are already implementing AI in areas such as financial and sales planning, chemical/contaminant monitoring, and back office paperwork automation.
Choking on change: Challenges of AI adoption
Having a full plate of options may seem tantalizing, but potential adopters face numerous challenges, chief among them cost. With margins already thin, F&B companies simply don’t have the deep pockets of companies like Google or Amazon when it comes to investing in AI.
Whether to build or buy is another critical decision. In an ideal world, F&B brands would build tightly integrated in-house technology that reflects the unique needs of their company. In the real world, the battle for AI talent is so severe that leading technology companies spend over $650 million annually to woo desirable candidates. Companies with established data analytics capabilities and a team of competent in-house developers may safely build their own AI platform. Those without such resources must instead seek out solutions and providers based on clearly defined needs, goals, and budgets.
Even for F&B companies that have found the perfect vendors, integrating a new AI system into existing technology stacks can be a headache, especially for large conglomerates with fragmented systems. Ken Wood, executive vice president of product management at logistics technology company Descartes, warns: “It’s painful to wire systems together — our customers tell us that consistently. The more vendors, the harder the project. The more systems you have to cobble together, the more expensive and longer time it takes.”
The final challenge remains the AI technology itself, which presents at least two issues for the industry. Without the right proprietary data, an F&B company may not be able to build machine learning models that perform. Matt Talbot, CEO of GoSpotCheck, describes this as “a huge obstacle without a cost-effective solution.” PepsiCo, Dannon, and Anheuser-Busch use GoSpotCheck’s AI-powered inventory software to maximize supply chain efficiency and provide business insights to sales reps.
Food and beverage companies are known to guard their secret recipes fiercely, but machine learning models should not be a mystery. Unfortunately, even with the right data, many AI solutions on the market work like black boxes. Without clarity and transparency into how algorithms are making decisions, F&B executives have a hard time determining whether a technology is truly adding value or how sustainable that value-add is.
CPG companies have experienced dismal growth in recent years. From 2013 to 2016, the industry grew less than 1.8 percent each year, on average. Complications of AI adoption aside, one fact is clear: F&B companies must invest in new innovations to cut costs, grow revenue, and stay current with consumer trends. Those who do may live to thrive another day. Those who don’t may find themselves replaced by tech-forward giants like Amazon.
This story originally appeared on TOPBOTS. Copyright 2018.
This story originally appeared on Www.topbots.com. Copyright 2018