The challenge: Help customers understand parcel size and postage requirements without relying entirely on manual measurement.
The opportunity
An Post handles an everyday task with real operational complexity. A parcel’s dimensions influence the options and cost of sending it, but customers may not have a ruler, a tape measure or a clear understanding of the size category.
Our approach
The work began with discovery and use-case exploration. We examined how computer vision might estimate dimensions from a smartphone camera and how the resulting information could fit into a customer journey.
The challenge involved much more than making a model recognise an object. A useful solution needed a clear interaction, sensible error handling, representative real-world tests and a credible integration path.
The engineering
Computer vision was explored through a proof of concept using parcel examples. Product and experience design considerations were developed alongside the technical approach, with contributions from programme partners.
What this illustrates
A promising AI use case needs both a technical proof and a practical user experience. Discovery, engineering and delivery planning need to work together from the outset.
