In recent years, much has been written about the Internet of Things and its potential to revolutionise the way businesses and public services operate. The vision, fast becoming reality, is for networks of connected sensors that gather data from factories, vehicles, hospitals, homes, shops and supply chains across the globe. It’s being hailed as the technology that will enable everything from smart cities that can, for example, optimise traffic flow, energy usage and signage, to systems that predict earthquakes.
You only have to look at the figures to see this is no mere hype. Earlier this year, market analysts IDC predicted that worldwide Internet of Things market spend will grow from $591.7 billion in 2014 to $1.3 trillion in 2019. By 2020, it’s estimated we will have 30-50bn Internet of Things endpoints worldwide.
In this article we explore what the ongoing impact of this will be on food safety. In particular, we examine the accepted wisdom that the Internet of Things, with its ability to monitor and control systems remotely, will help us go further towards guaranteeing food safety than ever before.
Could this really be the case?
Well yes, it could. But with something as complex as global food supply chains, it’s going to require a coordinated and collaborative vision that leaves no stone unturned, with every single component playing its part.
To fully understand why this is so, it’s worth recapping who and what is involved.
Food chain traceability
The Internet of Things is currently used most commonly in the food industry to track and trace the status of products through the production, processing and supply chain. At the most basic level, companies have been using barcodes and RFID tags for over 20 years to track food stuffs from their points of origin through to processing plants, transport, storage, distribution points and food retailers.
In more recent years, we’ve also seen the development of more sophisticated sensors that can monitor safety factors to a fine level of detail during food processing and logistics. It’s now relatively common for companies to install sensor networks that measure food dust particles, temperature or humidity in food manufacturing plants and transport containers, for example.
At the cutting edge of the field, we’re also seeing more advanced efforts such as the recently completed EU-funded MUSE-Tech project. This focused on applying insight derived from state-of-the-art sensing technologies (Photoacoustic Spectroscopy, Quasi Imaging UV-Vis Spectrometry and Distributed Temperature Sensing) to improve the way manufacturers handle raw and in-process materials.
The project developed a multisensor device that can react in real time to variations in raw material and processing conditions to optimise the quality and safety of processed foods. One strand of the project has focused on reducing the risk of developing the toxic compound acrylamide in starchy foods such as potato chips during cooking by specifying new guidelines for frying time and temperature.
Combined, the benefits of these approaches are many. For one thing, they help food companies comply with new legislation, such as the landmark US Food Safety Modernization Act (FSMA) signed into law by President Obama in 2011 and coming into force over the next few years. This specifically makes provision for improved food safety practices and greater accountability via traceability.
The other main advantage of end-to-end food safety monitoring is that it enables various parties involved in supply networks from producers, to logistics companies, to retailers to quickly identify points of origin and distribution if it’s discovered that food is contaminated. Previously these investigations could have taken weeks or months.
However, while all this capability is good for efficient food safety and identifying the sources of problems more quickly, it only takes us so far towards the true possibilities on offer. For that, we need to add Big Data to the mix.
Advances in Big Data analytics
As the Internet of Things develops in sophistication and scale, it is producing vast amounts of data that was previously unavailable. The question most industries are now asking is, how can we use all this information to improve the way we work?
This is the job of Big Data analytics, the rapidly maturing field of data science that can be used to mine exceptionally large datasets and uncover hidden patterns, unknown correlations, failure points, market trends, customer preferences and all sorts other useful business information.
In the food industry, datasets are being swelled by the vast amounts of data generated by the networks of monitoring sensors we’ve already described.
This is only just the start, however. What’s really interesting is the potential to find new correlations by analysing food safety data alongside sensor data from other scientific and environmental sources. This will create the possibility of delivering even more profound breakthroughs.
An example of how this can work is a current collaboration between Mars and IBM, which is focused on sequencing the DNA and RNA of bacteria that occur within foods across global supply chains. The goal of the project is to build up a genetic index of normal bacteria that occur in food.
This will help food and health officials to more easily determine anomalies in food samples that show the presence of harmful bacteria. It will give a better understanding of what causes contamination and then spread of foodborne diseases.
The project will apply analytics to the genomic data to look for new insights. It will also add relevant weather, transport and other contextual data gathered via the Internet of Things to see how this can help identify contamination breakouts early on.
When these elements come together the terabytes of genomic data from food samples, alongside vast amounts of data available from supply chain networks and other sensor networks we could see a new kind of analysis and insight that will ultimately take food safety to a new level.
Other sources of data relevant to food hygiene
Of course, there are also many other sources of data that can be analysed to advance understanding of food safety even further.
One such source is publicly available information that has previously been difficult to access and analyse but with the right approach could be put to good use. Chicago’s regulatory inspection program in the US has overcome this barrier by using Big Data predictive analytics techniques to analyse freely available data. This included using information such as food inspection reports, 311 service data, weather data and community and crime information to predict which restaurants are most likely to be in serious breach of food safety regulations.
When it analysed the data, Chicago found several key variables that could be used to predict the likelihood that a restaurant may have violated regulations. These variables included the establishment’s prior history of critical violations, through to the frequency of local burglaries, and even typical average temperatures for the location.
Since adopting the approach, inspectors have been able to uncover violations an average of 25% faster.
Another source of data is the information that comes from the many suppliers to the food industry. At Rentokil, this is something we know a lot about because we’ve recently started making our own contribution to the relevant food safety data pool through our PestConnect system.
PestConnect is built on a range of connected sensor-based devices that actively detect, capture or kill rodents and wirelessly transmit detection and capture information to a cloud server.
On one level, PestConnect is a system that helps us to remotely monitor pest activity, prevent customers having problems and control existing infestations. However, the data gathered also has great potential for applying Big Data techniques to identify sources and predict outbreaks.
To this end, we have now partnered with Google and PA Consulting to develop Big Data and predictive analytics capability with the aim of creating the next generation of pest control services.
With the right approach, this will ultimately help us ensure that our customers can take a much more proactive approach to managing food safety risks across global supply chains.
Multi-sector collaboration to improve food safety
On the one hand, the sheer volume and variety of data that is becoming available via the Internet of Things seems daunting. But on the other, it should really be seen as a massive opportunity â€“ especially if industry, governments and researchers can work together to share data and the insights the data could generate.
Thankfully, there are plenty of other organisations that seem to agree with this viewpoint. Nestle, for example, said only last year that that the huge amount of data collected by companies and regulatory authorities such as the UK Food Standards Agency (FSA) should be shared so that it can be â€˜minedâ€™ for information about emerging food safety problems.
This need to collaborate and share data and insight is also increasingly being recognised by industry groups like the Global Food Safety Initiative (GFSI), which wants to create an international multi-stakeholder platform for collaboration, knowledge exchange and networking along the supply chain.
Governing authorities are also playing their part. In 2015, for example, the European Commission’s Horizon2020 research programme called for proposals for a large-scale multinational, multi sector Internet of Things pilot project on smart farming and food security.
It stated that IoT technologies have the potential of helping European farming and food sector face important challenges for the future through real-time monitoring, better decision making, and improved operations management of the whole value chain, from farm to fork.
Significantly, the European Commission is encouraging the participation of all the potential contributors that play an active role in the agro-food chain, including farmers, machinery suppliers, food processors, retailers, wholesalers and of course the scientists and IoT technology suppliers working in the food sector. Results will also be used to inform EU policy on farming, food safety and food security.
The ability to combine multiple sources of data gathered in real time and extract new insights could have game-changing effects on food safety.
Predicting likely sources of contamination and food-borne disease outbreaks, Â detecting food safety threats as they happen and implementing control measures before threats can spread, will improve efficiencies and reduce losses throughout the supply chain. Most importantly, it will also improve food safety and security for all of us.