Nuru may have lifted cassava yield from 0.5 kg to 2–5 kg a plant

AI system

System: PlantVillage Nuru (offline TensorFlow convolutional neural network for cassava disease and pest symptoms on a smartphone)
Model: Not assessed
Release: Not assessed
Application developer: PlantVillage, Pennsylvania State University, with the International Institute of Tropical Agriculture (IITA)
Role: Not assessed
Configuration: Not assessed
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: A smartphone with Nuru supplied through a lead-farmer scheme (Self Help Africa, Ugunja Resource Center) with training. Use of the diagnosis to uproot diseased plants and select clean cuttings for replanting.
Human oversight: Not assessed
Relation to AI: Only the maker's account exists for this case. PlantVillage's post attributes the yield change, in the farmer's words, to Nuru's leaf-scanning diagnosis guiding removal of diseased plants and selection of clean cuttings. The independent evidence concerns the model's accuracy (Mrisho et al. 2020, developer co-authored) and another farmer's self-report in Business Insider, not Arisat. So the AI's share of her outcome cannot be assessed.
AI contribution: Not assessed

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: 2.0–5.0 kg per plant against 0.5 kg
Comparison: Assumed alternative: the same farmer's harvests before she used the Nuru app
Results: Assumed comparison. Reference: 0.5kg per Cassava plant. Result: between 2.0-5.0 kg per Cassava plant.
Additional value from AI: Not assessed
Limitations: Not assessed
Denominators: Not assessed
Materials for replication: Not assessed

Participants: Developer: PlantVillage, Pennsylvania State University, with the International Institute of Tropical Agriculture (IITA). Operator: Self Help Africa and Ugunja Resource Center (lead-farmer program in Busia County).
Jurisdictions: Benefit: Asing'e, Busia County, Kenya. Developer: United States. Operator: Kenya.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Conditions for replication: Symptoms are visible on leaves. Mrisho et al. report accuracy rising when six leaves per plant are scanned. The app's effect is separable from certified seed, spacing and market changes, which the post does not attempt.
Replication attempts: 1 follow-up entries. Benefit recurred: 1.
Later findings: 10 August 2022. By August 2022 Business Insider reported Nuru extended to eight more countries, with PlantVillage's David Hughes claiming cassava yield increases of 130% to 530%, and CGIAR's project page records 240 phones distributed to Bungoma, Kilifi and Tanzania in April 2021.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Assessed benefits: People.
Depth: People: 2.
Reasons: People: Arisat's reported rise from 0.5 kg to 2.0 to 5.0 kg per plant is a measurable improvement in her harvest. It comes only from the developer's blog, with no independent source.
Observed or intended: Not assessed
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: Independently reported

Kind: Not assessed
Related entries: Not assessed
Field: Agriculture / crop disease diagnosis. Smallholder cassava.
Purpose: Nuru is built to recognize cassava disease and pest symptoms on leaves offline and give management advice. Selecting clean cuttings from diagnosed plants is that intended use.
Dates: Date basis unassessed: 2020 (precision unassessed). Updated: 22 October 2026.
Stage: Not assessed
Version: SHA-256: 78586b05c4ff…

Peer-reviewed study Developer

Mrisho LM, Mbilinyi NA, Ndalahwa M, Ramcharan AM, Kehs AK, McCloskey PC, Murithi H, Hughes DP, Legg JP (2020). Accuracy of a Smartphone-Based Object Detection Model, PlantVillage Nuru, in Identifying the Foliar Symptoms of the Viral Diseases of Cassava-CMD and CBSD. Frontiers in Plant Science 11:590889 (PMC7775399; developer co-authored).
News report Independent author

Business Insider (2022). US state university developed an AI assistant for African farmers. 10 August 2022.
Press release Developer

PlantVillage (2020). Successful stories of cassava farmers in Busia County. PlantVillage blog, Pennsylvania State University, 23 March 2020 (developer-authored; linked from the CGIAR Big Data Inspire Challenge project page).
Press release Developer

CGIAR Platform for Big Data in Agriculture (undated). PlantVillage Nuru: Pest and disease monitoring using artificial intelligence. Inspire Challenge 2017 pilot and 2019 scale-up project page (partner-written; links the Arisat post as a success story).