To keep consumers satisfied, food processors – and their beverage counterparts – must continually innovate.
But they do so in an environment of growing regulatory burden, changing consumer preferences, and pressure to manage costs and effectively price products.1,2,3
Advancements in artificial intelligence (AI) and data analytics promise to help in areas from food safety and regulatory compliance to product innovation and margin expansion. The key is to integrate laboratories – R&D, testing, and quality assurance and quality control (QA/QC) manufacturing – into the enterprise’s overall digital ecosystem, enabling leaders to make better, faster decisions based on AI-fueled insights.
Food processing companies that deploy a single laboratory informatics platform to meet the needs of both research and manufacturing are in a better position to leverage AI, machine learning, and predictive analytics than those using siloed solutions. Frost & Sullivan, in its independent 2023 Frost Radar for Laboratory Information Management Systems report, argued that “the full potential of an ELN (electronic lab notebook) or LIMS (laboratory information management system) cannot be realized when used in isolation”.4 It advocated for a platform approach, including advanced analytics.
A comprehensive laboratory informatics platform serving the enterprise allows organizations to harmonize and standardize data fed into its data lake. This could include a flexible, agile ELN for researchers; a laboratory execution system (LES) to enforce manufacturing SOPs; a scientific data management system (SDMS) to automate data capture in a connected lab; and a LIMS as the backbone of data and sample management.
With integrated advanced analytics and capabilities for semantic search, all data, from R&D through to manufacturing, is accessible in a secure, roles-appropriate manner to be searched and analyzed for greater innovation, efficiency, and compliance.
The four key areas where an AI-powered LIMS platform can make a difference for food and beverage companies are:
1. Food safety
The World Health Organization (WHO) reports that almost one in 10 people fall ill every year from contaminated foods, resulting in over 420,000 deaths annually.5 Food processors, in addition to keeping consumers safe from the ill effects of contamination, seek to avoid the high cost – in dollars and reputation – of recalls.
AI can be leveraged in multiple ways to ensure the safety of the food supply. It can be trained using environmental monitoring data from air and surface sensors to understand the best conditions and alert to the presence of bacteria or other contaminants. AI can be applied to manufacturing specifications to finesse the ranges for safely producing finished products, alerting to indicators before they become failed batches. It can assist with shelf-life studies, for example, revealing trends in storage conditions over time. For supply chain issues affecting food safety, AI can be deployed in logistics planning.
Combine this with end-to-end traceability and companies can now stop a ‘problem’ before it enters the market, or be particularly precise when recalling products. Sample tracking that extends beyond the analyst and equipment level all the way to batch, lot, raw material, and finished product level can help producers rapidly identify products that need to be recalled, stopping small problems from becoming big problems.
2. Regulatory compliance
Many food and beverage companies are subject to multiple regulations across different markets and regions. In the US alone, regulatory agencies overseeing food production include the FDA, Department of Agriculture, and even the Consumer Product Safety Commission. Multi-jurisdictional compliance requirements can cause headaches for risk and compliance teams and make it difficult for lab teams to know whether they are checking the right boxes at the right times.
An AI platform that integrates with an organization’s existing data systems can save compliance teams significant time and frustration – and lower the company’s risk exposure – by allowing them to automatically enforce compliance within the system.
For example, AI can efficiently scan and analyze vast amounts of regulatory data, identifying changes, updates, and potential compliance gaps. By processing data on ingredients, production processes, and market trends, AI can pinpoint potential compliance risks and prioritize mitigation strategies.
Advanced analytics can support automation of document creation, review, and approval processes, reducing errors and speeding up compliance efforts. As well, it can generate compliance reports and facilitate audits, streamlining the process. It can assess the impact of regulatory changes on products, processes, and labeling, enabling timely adjustments. And it can support the development of compliance strategies and action plans to address regulatory requirements.
3. Product innovation and lifecycle extension
Ever-changing consumer demand – whether for plant-based foods, no-alcohol beverages, or more flavors of ice cream and chips – drives R&D departments. Having the right data at the right time is critical to product innovation and life cycle extension, which is why ease of access should be a key concern for companies investing in data systems.
Semantic search is a critical component of an AI-powered laboratory informatics platform; this capability allows users to retrieve scientifically relevant data based on context and meaning, not simply keyword matches.
Without powerful semantic search, many companies will end up duplicating past experiments, reasoning that it is easier to simply start over than spend the time determining what exactly was done in the past and what these past experiments yielded. Semantic search can save companies significant resources in R&D, and help these teams get to viable products much faster.
For example, semantic search has been used to determine how best to formulate a non-alcoholic beer. One company was able to leverage its existing recipe inventory, as well as data from external sources, to predict the optimal formula for an appealing new product. Using AI-powered research, exploiting the capabilities of semantic search, and using AI modelling and simulation, the company was able to significantly reduce the number of experiments needed to arrive at their desired taste profile.
While AI aids the R&D team and provides flexible solutions to test hypotheses, research data that is collected in a lab informatics platform can be stored in the organization’s data lake. Here it can become useful to downstream testing and manufacturing labs. Being able to capture a variety of information in flexible formats and enter this into the database alongside the rest of the company’s data is essential for getting a complete picture of a company’s products and processes.
4. Margin improvement
While it once may have been true that profit margins in food processing are generally lower than the average in other industries, the last several years have seen some improvement. The gross profit margin for the food processing industry in early 2024 was nearly 32%, above the total F&B market average of 25%. Net profit margin for food processing was 12% in 2024, up from 5% in 2019.3
Nonetheless, a history of tight margins means the food processing industry is a fiscally prudent one, balancing cost control as it seeks to meet consumer demand, regulatory requirements and food safety goals.
Any incremental improvement in margins can make a huge difference to a company’s bottom line. AI and advanced analytics can help these companies identify areas to improve efficiency and productivity, reduce costs, source lower-cost ingredients, adjust supply chains to reduce shipping requirements, and more. When conducted manually, these assessments can be expensive and time-consuming, while providing only a snapshot in time that can quickly become out-of-date.
An AI-powered informatics system, on the other hand, can continuously collect, organize, and analyze data, giving companies a constant overview of their products and operations. These insights allow for rapid improvements and the ability to make changes on the fly. The speed of data collection and AI-aided analysis in such systems accelerates the product development life cycle, and faster to market means faster to revenue.
What to look for in an AI/lab informatics platform
Centralized
The best analytics and AI solutions for food companies are integrated with LIMS, allowing for all of a company’s data to exist in a single platform. Companies need a centralized, centrally hosted, cloud-based solution – one that surfaces data from the company’s data lake house. With such a wide range of data feeding into the system, the insights the platform offers will be far more insightful.
Normalized
For LIMS to dynamically and easily interact with a larger data lake house, the setup requires a commercially available relational database, such as Microsoft SQL Server and Oracle. The database should be normalized, allowing teams to query it to get the data they need. The platform should store data in a non-proprietary manner, allowing users to extract and read the data without an additional tool to interpret it.
Configurable
To facilitate the type of analytics and AI that drives change, food and beverage companies need a truly configurable system that allows them to make decisions, rather than one that enforces rules upon them that may not always apply. Individual companies should still be able to determine how they track, record, and store their data.
Any analytics and AI solution should allow for systems to be dynamic down to the user, job type, or user group level, surfacing only what is relevant for particular users. With multiple departments and team members using the same platform, configuration is critical.
One large multinational food and beverage producer operates a single laboratory informatics platform across multiple R&D sites and hundreds of manufacturing sites around the globe. This AI-powered system is tailored to meet the needs of each of the company’s relevant teams, enabling the organization to start leveraging AI for product life cycle decision-making.
Authors: Dan Call, Strategic Account Executive, LabVantage Solutions and Alan Marcus, Chief Growth Officer, LabVantage Solutions
References
1. Foley. What Food & Beverage Companies Need to Know About the U.S. Consumer Product Safety Commission.
2. Green Hasson Janks. Food Innovation the put and pull of consumer demand.
3. Investopedia. Profit Margins for the Food and Beverage Sector.
5. World Health Organization. Food safety.