How can AI revolutionize traditional horizon scanning?
There are different ways that food safety teams can stay informed about incidents and emerging risks that could be critical to their supply chain.
Welcome to the world of food risk monitoring tools – also widely known as horizon scanning and foresight tools
For many years, these terms were used to describe online services that scanned a large number of information sources in order to alert professionals as soon as an announcement of interest was published. Prominent services widely used for this purpose include FoodTrack Inc., the IFS Trend Risk Monitor, and FERA’s HorizonScan.
In an era of AI and predictive technologies for food risk prevention, available options for performing risk monitoring across the food supply chain have become even more extensive and powerful.
Let me walk you through the main ways in which food companies address this need.
Option 1: Rely upon suppliers to inform them about a critical incident
This is the most traditional way to be informed about an incident that concerns their suppliers. However that rests on the expectation that someone will pick up the phone (or drop a line), as soon as an incident occurs. They may not be proactively on the lookout for particular incidents that could be critical to their customers’ supply chain.
This approach assumes that food companies fully trust their suppliers and that, if something happens, they will be notified on time.
Option 2: Regularly monitor 2 or 3 official sources of information in the markets of interest
This means devoting less than a couple of hours every week to visit reputable and official websites (such as RASFF) to check if something new has been announced. This type of routine keeps food safety professionals up to date with credible, verified information. This method is particularly useful if their geographical and jurisdictional scope of interest is covered by these specific official authorities.
This approach assumes that official authorities have everything that matters to a food company on their websites.
Option 3: Systematically devote time to research numerous online sources about emerging issues and hazards
Food professionals doing their due diligence comb through targeted resources to be more proactive and extensive in their information search. Apart from the 2 or 3 major global authorities, they may also bookmark a number of other information sources that they systematically research. Very often, these include:
- Inspection reports announced by local authorities near to their critical suppliers
- Websites of official authorities in each and every country in which their products are being distributed
- Results of border inspections or residue monitoring programs in these countries
- Trusted food safety news sites
This approach can really keep brands up to speed with what is happening. It assumes that they have the people and time to allocate to this form of systematic desktop research.
Option 4: Use a third-party service that provides notifications on announced incidents as well as predictions for expected ones
Whether they choose one of the traditional horizon scanning tools or go for a modern AI-powered dashboard, food supply chain companies can find plenty of third-party solutions available – and for every budget.
Signing up for an external service effectively outsources the manual work of scanning, processing, combining and extracting meaning out of dozens or hundreds of information sources. Third-party services save lots of time and resources, especially if the selected tool also predicts emerging risks and upcoming incident trends so that it gives them a glimpse into the future.
This approach assumes that they have explored several available options and have chosen a solution that fully covers their company’s needs.
Option 5: Develop an internal software that monitors a variety of online sources, provides with alerts & notifications, and implements AI algorithms that predict upcoming issues
The preferred solution for large food manufacturers that have the appropriate resources, senior management buy-in, and a digital transformation agenda.
Software development focuses only on the modules and features that are important to their company’s priorities (although benchmarking against others in the industry usually helps). It reduces the risk of relying upon third-party suppliers for a critical operation.
This approach assumes that, apart from the required resources, the company demonstrates patience and persistence. In my experience, organizational data science capabilities and solutions grow and mature with time.
So, looking at all these different ways of implementing a food risk monitoring strategy in a food company, where can AI & predictive analytics make a difference?
(1) Horizon scanning & alert investigation
- Website crawling software can automate data gathering and processing, while monitoring updates in near-real time.
- Text mining algorithms can understand which products, ingredients and suppliers are involved in a reported incident.
- Deduplication algorithms ensure that each report is unique, aggregating multiple alerts into one, more complete data record.
(2) Identifying rapidly emerging issues
- Time-series forecasting can estimate how many incidents per product or ingredient category should be expected in the weeks or months to come.
- Prediction models can estimate whether specific risks have a high or low probability of occurring.
These are only a few of the many ways in which AI-powered technologies can significantly expand the capabilities of food risk monitoring systems.
Want to learn more about where your company stands in terms of food risk monitoring? Take this free online self-assessment test.