--- title: The cold chain, illuminated slug: cold-chain summary: Led design for a smart-pallet venture that makes spoilage visible from farm to fork. Surfaced a $50m+ global opportunity, converted five paying pilots in the first twelve months, and co-created self-serve onboarding with fresh produce growers in remote New Zealand. url: https://charlieholland.com/#work/cold-chain --- # The cold chain, illuminated **Client:** Agritech, New Zealand · **Role:** Lead Product Designer · **Year:** 2024 to 2025. Led design for a smart-pallet venture that makes spoilage visible from farm to fork: a tracked pallet whose embedded device reports position and temperature as it moves through the chain, and the software that turns that stream into a spoilage story people can act on. The work ran from ecosystem discovery through to the onboarding that let customers start without anyone standing beside them. Headline outcomes: a $50m+ global opportunity identified through the discovery work, and five paying pilots signed in the venture's first twelve months. Onboarding was co-created with fresh produce growers in remote New Zealand and designed so customers could get started without any hand-holding. ## Overview: a smart pallet that shows you where the cold chain breaks Fresh produce loses value silently. Between the paddock and the shelf it changes hands half a dozen times, and every hand-off is a chance for temperature to drift with nobody watching. By the time spoiled produce is visible at the shelf, the cause is days and several hand-offs behind it. The venture's answer was to put the tracker inside the pallet itself, so the thing that carries the produce is also the thing that witnesses its journey. I led design for the venture end to end: the ecosystem discovery, the journey definition, the pilot focus, and the self-serve onboarding. ## Challenge: spoilage happens where nobody is looking A carton that froze in transit and a carton that sat on hot tarmac look identical on arrival, and nobody in the chain can say which leg did the damage. I mapped the chain from producer to retail shelf and marked the fourteen places it hurt. Every party on it was carrying cost from a failure they could not see and could not prove. The discovery mapped the full ecosystem a pallet of produce moves through: producers, supplier transport, packhouses, distribution centres, retail carriers, retail stores, Retail HQ governing both ends, and the pallet-services loop of carriers and service centres that inspects and reissues pallets. Painpoints surfaced at every hand-off: - Producers struggled to evaluate and select third-party logistics providers on performance, and were asked to provide evidence of journey temperature for rejected produce they had no way of capturing. - Packhouses had no evidence for claiming damaged produce against transporters. - In transit, human error disrupted temperature stability and product quality. - Distribution centres faced manual, time-consuming quality checks across large volumes, varying temperature measurements on arrival due to RF gun inaccuracy, inaccurate or delayed insights from data loggers, and bore the cost when quality issues were not detected on arrival. - Retail HQ struggled to evaluate and manage suppliers and the reliability of 3PLs. - Retail carriers saw frozen produce and impact damage affect quality, and temperature fluctuations compromise shelf life. - Stores received degraded produce when not all cartons were checked at the DC, and could not grade local produce within their cost and time windows. The finding that shaped everything after it: blame travels with the pallet, but data does not. Rejected produce triggered claims and counter-claims with no evidence on either side, and the gaps between legs of the journey belonged to no one. That gap between accountability and information is where the opportunity sizing started. ## Approach: walk the journey before designing a screen No product surface existed yet, so the journey had to be the artefact: what happens to a pallet at each stage, where temperature risk concentrates, and which moments the tracker had to witness for its data to be worth paying for. The journey breaks into four stages, each with its own pain and its own opportunities for a tracked pallet: 1. Producer and packhouse: difficulty evaluating supplier performance, lack of evidence for damage claims, and challenges with temperature monitoring. Opportunities: supplier performance analytics, a journey temperature evidence system, and product quality certification. 2. Transport: human errors disrupting temperature stability, no evidence for claiming damages, and temperature fluctuations affecting quality. Opportunities: real-time temperature monitoring, digital journey logs, and transporter performance analytics. 3. Distribution: manual, time-consuming quality checks, costs borne for undetected quality issues, varying temperature measurements, and the challenge of managing large volumes. Opportunities: a guided QC assistant, an integrated QC platform, borderline checks, on-arrival quality flags, and automated quality monitoring. 4. Retail: degraded products reaching stores, no local produce grading, limited time for quality checks, and shelf-life uncertainty. Opportunities: estimated shelf-life tracking, a quality verification system, automated monitoring, and real-time notifications. The pilot focus landed on the first two stages, producers, packhouses and transport, where the evidence gap was sharpest. That made growers the wedge: if the venture could not win the people who load the first pallet, nothing downstream would ever see the data. ### Onboarding you can finish in a paddock Growers in remote New Zealand had a hard constraint: no one was coming to set this up for them. So we co-created the onboarding with fresh-produce growers themselves, and designed it fully self-serve. The co-creation sessions prioritised every onboarding question by how critical it was and how well we understood it. Five became the pilot onboarding, in priority order: data and privacy, ordering, the Ultra Device (the tracker itself), smart pallet and product association, and the Brix scan app. Organisation admin, billing and costing, and storage were deliberately deferred. ## Results: five pilots paying inside the first year What convinced them was visibility they had never had. A pallet that reports where it is and what temperature it is, and a rule that turns a sustained breach into a named leg and a named custodian. The discovery work sized the same problem globally at a $50m+ opportunity, which gave the venture its case for scaling beyond the pilot region. The onboarding was designed to be finished in the field without a site visit, and the pilot was the test of that service model: a customer base too remote to visit can still be a customer base. ## What was hard Designing for hands that are full. The product gets used on loading docks and in paddocks: gloves, glare, patchy signal, nowhere to put anything down. Every flow had to survive being interrupted halfway through and picked up again without loss. A service, not a screen. The design surface was the whole hand-off between growers, transporters and retailers, with pallet hardware in the middle of it. Most of the important decisions were about who sees what evidence and when, not about interface. Zero-to-one ambiguity. We were sizing a global market and shipping pilots with the same small team at the same time. Deciding what the pilot had to prove, and what could stay unknown, was the discipline that kept it moving.