Alireza Ehsani

Data-driven breeding of resource efficient cattle 

Feed efficiency is an important production, economic and environmental factor in cattle production and accounts for as much as 50% of total costs. The dairy industry alone is worth £188bn globally, £56bn in Europe and £3.5bn in the UK. The primary objective of this project is to benefit from data-driven innovation tools to innovate a cost- effective and accelerated breeding scheme for feed efficiency in the dairy cattle industry. Efficient cows in terms of feed conversion ratio to milk production emit less greenhouse gases, including methane (CH4), which contribute to global warming. The proposed work has the potential to deliver about 1% genetic improvement in feed efficiency per year with a value of ~£2m per year for the UK.

Owing to smart sensors and a range of computer applications combined with high-tech engineering enables the dairy industry to collect a huge amount of data. This project will enable breeding of resource efficient cattle by developing new systems for feed intake recording, statistical modelling of the new and existing data, and genomic evaluation that are informative, scalable and cost effective. The main premise of the project will be to assess how resources for phenotyping are spent in a dairy breeding programme for a range of traits and how to reallocate these resources so that improvement can be more balanced.

The project team is collaborating with the breeding organisation of Norwegian Red dairy cattle named GENO.

To deliver the aims of the project we will work with an interdisciplinary team from university and the industry partner in the following order: 

  • The breeding system data from the industry partner will be assessed to optimise the breeding plan.
  • We will develop an in-silico modelling system of the partner’s cattle breeding programme.
  • We will use the modelling system to design a feed intake recording system that generates maximal amount of information per investment we will achieve this with a statistical modelling of new expensive feed intake data, new cheaper feed intake indicators and existing milk recording data (millions of milk samples with hundreds of mid-infrared spectrometry wavelengths) to predict feed intake for all animals.
  • We will deploy the proposed data system in the partner’s breeding programme and use it for genomic selection to breed future generations of resource efficient cattle.

This TRAIN@Ed project has received industry funding from GENO and funding from the DDI programme, the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 801215.

For more information visit Edinburgh Research Explorer Profile.

Breeding planet-friendly cattle: COP26 article

Contact email


Our Women in Data campaign

Image of Natalie Duffield

Natalie Duffield

CEO of InTechnology Smart Cities Ltd

Photo of Hoppers Society Members

University of Edinburgh Hoppers Society

Image of Samantha Rhynas

Samantha Rhynas

Head of Data at Effini & Girl Geek Scotland Leadership Team & PyData Edinburgh Organiser