Harvest event detection is an existing Digital Harvest capability developed to provide measured field-level harvest dates. It allows crop age, growth months, and crop-cycle timing to be calculated from observed field activity rather than relying only on self-reported harvest dates.
After reflecting on conversations with KSB about cane poaching, we explored whether this capability could be applied in that context.
This sample identifies likely harvest windows across 172 sugarcane fields and highlights harvests that occurred outside the normal milling period. Harvests outside the milling window may warrant additional review because the timing may fall outside standard mill operations, contract expectations, or regulatory oversight processes.
One potential application is to compare detected harvest dates with mill delivery records. If a field was harvested but there is no corresponding delivery to the contracted mill within the expected post-harvest period, that field becomes a candidate for further investigation.
This report does not conclude that cane poaching occurred. It identifies where harvest activity, timing, and delivery records may not align, allowing mills and regulators to focus investigation on the fields where discrepancies exist.
This sample is built on fields from a separate engagement, not Kenya. Applying this approach within your organization would require registered outgrower boundaries and mill delivery records.