How to Automate Croptracker

The spray record is the only proof the block was clean

An apple leaves the orchard with a history attached to it. Which block it came off, which crew picked it, what went on that block and how many days before the pick, which bin it went into and which lot code that bin became. Croptracker is where a grower keeps that history.

Dragonfly IT has been building it since 2006, for fruit and vegetable growers, packers and the co-operatives that market for them. It is not one thing. Field mapping and planting, spray record keeping, crew activity and labour tracking with piece-rate calculation, harvest yield records, storage records, packing and shipping traceability records, quality control, and the GAP reports that pull it together at audit time.

The reason it matters is not the reporting. It is that a fruit crop is a chain of custody. When a buyer calls about a lot, or an auditor asks who applied what on which block, the answer has to already exist. Nobody reconstructs a season in October.

Harvest runs on numbers that arrive by phone call

Growing is not the part that eats the week in September.

Bins come in and somebody counts them. The picking crews' hours get collected off tally sheets and reconciled against what the piece rate says they earned. Then the fruit goes to the packhouse and the farm stops knowing anything about it for a while. Weeks later a packout statement comes back: what graded out, what dropped to juice, what the pool actually paid. Somebody reads it, works out which blocks did badly, and writes that down somewhere.

The buyer side has the same shape. Purchase orders sit on a retailer's vendor portal. Delivery appointments get booked on a different one. And on a clear night in April, somebody is checking a forecast on their phone at eleven, deciding whether to wake anyone and start the wind machines. Every one of those is a screen a person has to open, at a time a person has to remember.

Croptracker

Packout traced back to the block Grade returns read off the packhouse portal the day they post, then split back to the blocks and pick dates that produced them.

Frost calls that name blocks The overnight forecast set against the blocks your record says are at their most vulnerable stage, so the decision to run the wind machines is specific.

Audit status watched all year Certificate dates and renewal status checked on the certifier's own site each month, rather than the week a buyer asks to see them.

Croptracker calculates the interval it was told about

The spray module is the strongest argument for the software, and a grower still keeping a paper book should hear it first.

Log a spray on the phone in the row and it captures the chemical, the rate, the block, the weather and who ran the sprayer. From its chemical database it works out the pre-harvest and re-entry intervals, and the field map goes colour-coded: restricted blocks and clear blocks, updating as the hours pass. Chemical inventory deducts as sprays are logged, so a low-stock warning arrives before the shortage does. At audit, the GAP report comes out in the format the auditor expects, carrying the product, the EPA registration number, the date, the rate, the block, the weather and the applicator.

That is a compliance problem genuinely solved. But look at what the calculation rests on: an interval Croptracker was told, off a label already in its database, about a spray somebody remembered to log. It has no way to check whether a registration changed this season. It cannot open the packhouse portal to see how the fruit actually graded. It cannot read tonight's forecast.

A frost warning that reaches the crew leader by itself

Everything here starts in the orchard record, which makes the outward direction the interesting one.

Picture the forecast checked every evening through bloom against the blocks the record says are at their most vulnerable stage, so the decision to start the wind machines arrives as a message naming blocks rather than a number on a weather app. Picture the packout statement read off the packhouse portal the morning it posts and split straight back to the blocks and pick dates that produced it, so the block that graded badly is named while the reason is still fresh.

Picture your certification checked on the certifier's own site each month: the audit date, the certificate status, whether the renewal went through, so nobody finds an expired certificate the week a buyer asks to see one. Or the retailer's vendor portal read each morning for new purchase orders and delivery appointments, set against what is actually sitting in cold storage. Or a product's registration status checked on the regulator's own database before it goes in the tank rather than after.

Certificates and packouts read while the crew sleeps

Every one of those sits on a website somebody has to sign into and read. That is the whole obstacle. Not difficulty, just a person and a browser at a particular hour.

WebRun is that browser. It opens the packhouse portal, the certifier's site, the retailer's vendor page and the forecast, signed in as you, on the schedule you choose, and puts what it finds where you will see it. No API, no export button, nothing for the packhouse to build first.

The reading runs unattended. Anything that changes a spray record, or that sends a message to a buyer, a certifier or a crew leader, is drafted and waits for a person. A spray record is a legal document and a delivery slot is a promise, so those keep a human on them. The cards below are the orchard jobs growers tend to start with.