The Tipping Point for Machine Learning in Agriculture

The Tipping Point for Machine Learning in Agriculture

At the PrecisionAg Innovation Series earlier this year, retailers and growers both expressed the need to gather data and the frustration they felt with the lack of services available to refine it into actionable insights.

Advertisement

The precision ag landscape is changing. The number of sensors on machines is increasing. Agricultural and data science experts are working together, leading us to a point where accurate prognostics can be shared with the right person, in response to localized environmental conditions. The investments which industry giants have made to access this data signifies its value. BASF’s acquisition of ZedX, Monsanto/Climate Corp.’s acquisition of HydroBio, and Dupont Pioneer’s acquisition of Granular have all been announced since May of this year. The last of which came with a price tag of $300M. These acquisitions have led to the consolidation of data sources, which further enables accurate prognostication.

The realization of data is going to begin with the machines. Other industries including rail, energy, mining, aviation, and construction have already realized enormous benefits from using data science and machine learning to increase reliability, productivity, and the safer operations of their equipment. These industries all share something in common with agriculture; they all require massive, expensive machines to run productively for extended periods of time. Each of these industries fights against machine downtime, repair costs, and trying to optimize which components to replace at the right time to maximize their utility.

The railroad industry is a particularly analogous to ag, and in rail these same challenges were successfully met head on with machine learning and predictive maintenance. In just one year, Uptake Rail performed a test with the data of a prominent North American Class I railroad, and was able prove the ability to save $160K per locomotive per year through reduced asset downtime, fuel consumption, and the time period it takes to build a train (i.e., assign a healthy locomotive to a mission with confidence). In agriculture, we can easily leverage the following tools that have worked so effectively for rail:

Asset Management

  • How healthy are my machines?
  • Where are my machines?
  • How productive are my machines?

Dealer Solutions

  • When can I expect maintenance scheduling?
  • How can I accurately predict the amount of inventory to order?

Workflow Optimization

  • How efficiently are my machines being deployed?
  • Can we further optimize grain carts, fill ups, and fleet management to make sure planters, combines, and sprayers are all running as efficiently as possible?

Uptake’s single-code base platform leverages machine learning to transfer its value from one industry to the next. We have worked with progressive industry leaders in rail, energy, mining, aviation, and construction.  We are excited to see what’s possible in agriculture; who else is excited to learn the same?

Leave a Reply

[…] Au final, le Machine Learning est une technique qui permet d’instruire les ordinateurs pour qu’ils s’adaptent aux changements apportés aux données, produites en permanence et de manière illimitée. L’analyse faite par des ordinateurs à l’avantage de la rapidité, la précision et l’objectivité comparé à celles faites par l’homme. Autant d’atouts qui en font une technique incontournable, qui atteint désormais un seuil critique. […]