Cutting-Edge Data-Driven Predictive Models
Harnessing Artificial Intelligence for Precise Poultry Production Forecasts
Data-Driven Forecasts Without Historical Input
Intelia has put the power of artificial intelligence, namely machine and deep learning, to create poultry-specific predictive models that rely 100% on current data to forecast the prediction. These autonomous models are so evolved that they do not require growers to supply their own historical data to adjust the models. In most cases, they will self correct automatically.
We currently have 4 autonomous predictive models
14-day continuous bird weight prediction
The continuous 14-day prediction makes it easy to spot if a flock is running behind vs what is expected. When a target weight is provided, the prediction is used to estimate when it will be reached. Projected over 14 days, even a small weight discrepancy today will be seen more clearly, enabling growers to detect issues 36 hours earlier than relying on average weight alone.
The red dotted line indicates expectations, while the gap between the red and blue lines reflects the flock’s performance. The white section represents the historical performance of the current flock, while the blue portion corresponds to the anticipated performance for the upcoming 14 days.
Benefits
- Monitor target weight vs expected date of harvest
- Make adjustments on flock management to influence the growth as desired
- Monitor the impact of specific feed formulation or additive
- Forecast expected revenues from a specific flock.
Time to target weight
The capability to consistently monitor and predict the moment when a flock attains its target average weight is invaluable for efficient coordination with the processing plant. Moreover, this data serves as the foundation for creating an optimized harvest schedule, maximizing the average live weight at the plant. However, recognizing that average weight is just one facet of the equation, our system provides additional insights by revealing the flock’s uniformity and standard deviation. These metrics are essential for streamlining planning processes and optimizing processing lines, ensuring overall efficiency in poultry production.
Time to empty feed bins, adjusted in real-time with mortality
By tracking feed consumption and current population within a specific house, Compass offers valuable support to your feed mill operations by providing precise predictions for when bins will run out of feed in any connected facility. Configurable notifications for low feed levels and alerts for empty bins ensure swift responses, guaranteeing that your birds never face feed shortages.
Benefits
- Prevent over-delivery and resulting feed pickup/credit
- Minimize unnecessary handling of the feed
- Reduce the use of picked-up feed in other houses, which can lower nutritional value due to a higher level of fines.
Feed required to bring birds to target weight
This predictive model is useful to optimize feed delivery in the final days of a flock. Adjusted with feed consumption and population, Compass will calculate the precise amount of feed needed to reach the target weight based on actual feed conversion and/or feed consumption and adjusted with actual mortality.
Benefits
- Prevent feed outages
- Eliminate out-of-sequence deliveries and regroup resupply of neighbouring farms
- Optimize feed delivery costs by ensuring trucks are maximized and routes are well planned
- Eliminate non-value-added task of checking feed levels with a more accurate way of measuring inventory.
Put the power of your data to work today!