a few days ago, the “ai pig breeding” project of the group achieved key results, and the “intelligent data collection method based on computer vision image recognition” won a national patent, paving the way for true ai pig breeding.
mastering the core technology can truly realize “pig breeding based on data”
in the era of african swine fever, the speed of industry integration will not stop because of african swine fever. on the contrary, it will become a watershed of breeding enterprises, where the strong enterprises are stronger and the weak enterprises are out.
in fact, there are indeed many technical limitations in the human-dominated breeding industry, but it can also be understood as an opportunity for the rise of new giants-provided that you can seize the opportunity and change quickly.
nowadays, with the intensification of industry integration, especially the huge impact of african swine fever, ai breeding has become a hot word in the industry. however, there are more “expectations” and “prospects”, and there are few “influential achievements” that can be truly implemented.
data, computing power and algorithm are three elements recognized by the industry as the basis of artificial intelligence. however, the digitalization of china’s breeding industry and the application of ai technology started slightly later than developed countries, and one of the most important factors is that domestic breeding enterprises have limited involvement and little contribution in the data collection stage, which is the basic link of ai.
due to the lack of ai solutions related to data acquisition, processing and analysis, many artificial intelligence projects are unable to make breakthroughs in the field of pig farm data, and the backwardness of data collection methods has led to the failure of a large number of breeding companies to truly implement “data-driven” approaches.
liberating pig farmers can truly realize “scientific pig breeding”
if you are a pig farmer, then you must be familiar with the following situation:
1. data acquisition methods are outdated and inefficient. at present, most of the farm data in china are recorded through manual or semi-manual means. the production data are either manually recorded into various paper forms by pig house manager, converted into electronic documents and then entered or imported into the production management system, or entered one by one at the production site through mobile app. some breeding houses need to fill in more than 10 forms--data collection becomes a heavy burden of pig farmers.
2. the data quality varies. most enterprises have serious data missing and poor timeliness, and the accuracy of data is even more impossible to guarantee, which brings great uncertainty to subsequent scientific decision-making.
how to release the pressure of data management in the process of pig breeding and ensure timely and accurate data collection is a problem that dekon’s “ai pig breeding” project must solve.
through in-depth research on the acquisition, storage and analysis of pig breeding data, the project team independently developed intelligent data acquisition and business data conversion technology based on business mapping cards and computer vision. with the support of such technology, the content of the new breeding pig management card is no longer the traditional text, but the standardized and quantified business data. operators only need to fill in the map according to simple rules, and each rule is bound to a business. the technology now has covered the data of breeding pigs in the whole production process, so that the data collection is no longer isolated, but forms a good service data flow.
the benefits brought by technical breakthroughs go far beyond that.
“the previous breeding enterprises are human-oriented, and both information collection and pig mating need to be done manually, which not only has low efficiency and unstable accuracy, but more importantly, it is not the most effective way to play the role of “people”.” zhang yong, a member of the project team and vice general manager of zhenghu technology, said that: “to pursue greater and faster development, we need more innovative talents, and the most important significance of artificial intelligence is to liberate these new pig farmers from tedious and repetitive work.”
