PEAT

Deforestation and soil degradation continue to pose a significant threat to food security and have a negative impact on the economy. The USDA estimates the annual cost of soil erosion at about $44 billion.

Berlin-based agricultural technology company PEAT has launched Plantix, a deep learning app that claims to identify potential soil defects and nutrient deficiencies. The analysis is performed by software algorithms that correlate individual foliage samples with specific soil defects, crop pests and diseases.

The image recognition app identifies potential defects through images taken with the user’s smartphone. Users are then provided with soil remediation techniques, tips and other possible solutions, according to a short video shown below:

The company claims its software can quickly achieve pattern recognition with up to 95 percent estimated accuracy. PEAT recently announced that its international customer base has reached more than 500,000. The company acknowledges its partners, and customer quotes are on its website, but no specific case studies appear to be available.

Without specifications on the size of client farms, we cannot confirm whether the Plantix application poses significant limitations for larger farms. Competitor CropDiagnosis appears to be following a similar model for its application.