‘The project is a collaboration between meat producer Vion and animal welfare organisations De Dierenbescherming and Eyes on Animals. Their goal is to improve the monitoring in slaughterhouses by means of innovative video software. This can help to reveal potential malpractice’, Vincent says. ‘Dutch slaughterhouses have increased the use of video systems in the last few years. However, the cameras are on 24/7, so manual processing of the footage is nearly impossible. In practice, this meant that random selections were made, and potential malpractice often remained hidden.’

Vincent’s team was involved in the set-up of a pilot project to assess whether AI-based software could be a solution. ‘We started with the unloading of pigs from trucks. In this situation, there is quite a lot of human-animal interaction, and more risk of deviations in the animal welfare protocol. Moreover, it offers an accurate overview of all pigs arriving at the slaughterhouse’, Vincent explains. ‘We have analysed thousands of hours of footage to determine the relevant information. Where are the animals? Where are the humans? Where are the objects? What is their interaction like? Which actions are according to the protocol, which are not? If a pig lags behind, that could be the indication of an injury.’

'The animal welfare organisations have helped us with their expertise in animal welfare, and Vion has been very open from the start - sharing footage and showing us around in a slaughterhouse. A collaboration between these parties may seem very unlikely, but obviously there is a strong common denominator: monitoring potential animal suffering. All parties involved recognise that improvement can only come from collaboration. And the result is impressive’, Vincent says passionately. ‘By feeding the system with various examples of unwanted situations, it learns how to recognise patterns. And this learning process never stops. If the system runs into deviant and unwanted situations in the footage, a notification pops up in the dashboard, so employees are able to assess it. The feedback on the observations is subsequently used to further improve the model.’

‘This is such an interesting project’, Vincent says. ‘It is our mission with the Deloitte Impact Foundation to contribute to a better society. That includes the reduction of suffering of all living beings, including animals. Moreover, with the theme The Future of Food we are focused on the modernisation of the food supply. That includes a better distribution, waste reduction and the reduction of negative side effects during the production, such as human and animal exploitation. To be honest, I haven’t always been aware of this, but this project has been such an eye-opener. I have learnt so much, both on a professional and on a personal level. Our team is an interesting mix of disciplines. I have worked with some of the team members on the development of the AI component. But the privacy and security issues of the system are pivotal as well, as is the question how we can convince employees to use the new system. Together we work on a common goal. And even though, due to COVID-19, we have largely been working from home for more than a year, all of us are so passionate about this project that it feels as if we are working in the same room.’ This initiative is supported with hours from the Deloitte Impact Foundation.

CHOOSE YOUR IMPACT You share your expertise with the 1500 colleagues of Deloitte The Netherlands Technology & Transformation department. In this department, innovation and personal development are very important. View all Technology & Transformation vacancies [https://careersatdeloitte.com/vacancies/technology-transformation] and connect your future to Deloitte. View all Technology & Transformation vacancies [/vacancies/consulting]