Project facts

  • Challenge
    Detecting potential quality of life issues by physically examining thousands of pigs every single day.
  • Solution
    An automated physical inspection solution that uses AI to analyse each pig’s tail length – a key indicator of life quality.

Results

  • Supports animal welfare protection efforts

  • Enabled simple supplier benchmarking

  • Replicable solution that can be deployed on any processing line

The challenge

Supporting ethical pig sourcing

For meat processors, consistently sourcing ethically bred and raised animals is an extremely difficult task. Our client makes an active contribution to improving animal welfare in all parts of the chain in which they are active. But with low visibility into supplier breeding and rearing activities, and a limited view of the conditions animals are raised in, it’s very difficult for any processor to gain 100 percent confidence in the meat they buy.

As a result, many processors simply feel that improving animal welfare is beyond their control. But, our client recently challenged that idea, instead asking “what can we do to better monitor, benchmark and improve the quality of life for the pigs we acquire?”

With thousands of animals passing through its processing facilities every day, the company knew that whatever solution they created to help support that goal, it would need to be powered by intelligent, automated technology. That’s when they engaged Royal HaskoningDHV Digital.

The approach

Enabling automated visual inspection

Detecting the physical signs of poor animal welfare can be very difficult. Fortunately, with pigs, there’s a common physical indicator of their quality of life – their tail length. If a tail is short – or removed entirely – it’s typically a sign of the pig being raised in poor conditions.

With that physical differentiator determined, we worked with the client’s domain experts to begin planning a monitoring solution that could automate tail length inspection, and provide valuable insights back to the business with no human intervention.

After identifying the ideal physical location for cameras, and creating AI models to process the visual information they generate, we deployed the solution and began testing it in operation for the first time.

The solution was designed with demanding production environments in mind, ensuring it could be implemented according to the highest hygiene and robustness standards set out by IP69K. Plus, the AI powering the solution also ensured that it wouldn’t be disrupted by common environmental changes like shifts in light level.

The impact

Enabling improved animal wellbeing

The client is continuously engaging in activities to help improve animal wellbeing amongst its suppliers and ensure that suppliers are upholding their quality of life commitments.

Today, with our AI-powered solution in place, our client be able to monitor the tail length over time. When the average length increases due to the interventions they’ve done, this will have a very significant impact on animal wellbeing. Especially since our client processes about 15 million pigs per year.

Empowered with verifiable data to support their efforts, they can see the real impact of their animal wellbeing projects and programmes, benchmark suppliers against one another, and recognise when there’s a need to take further direct action with a specific supplier.

The solution was designed with production environments in mind, ensuring it could be implemented according to the hygiene and robustness standards set out by IP69K.