The added value of a distributed sensor network

Field research · Langhe, Monferrato, Turin Hills

One weather station is not enough

Simple and common questions, but answering them requires detailed knowledge — you need to know every zone and characteristic of the vineyard. Not a representative point. Not data from the nearest weather station. Data from the vineyard itself, along the rows, inside the canopy. None of this is simple: deploying a distributed sensor network in a real viticultural context is technically demanding, but the enormous advances made in recent years now make this — and much more — possible.

A weather station measures what happens at a single point, but the natural complexity of a vineyard cannot be condensed into a single point. Meso-climatic and micro-climatic variability within a vineyard is well documented in the scientific literature. Studies conducted on Nebbiolo vineyards show how exposure, elevation and canopy vigour determine significant differences in air temperature, canopy humidity and solar radiation, even within the same plot. Micro-climatic characteristics, as highlighted by precision viticulture research, can differ even along the same row.

METER Group, a provider of instrumentation for agronomic research, is explicit on this point: between the meteorological conditions measured at a specific point in the vineyard and those provided by stations located even a few hundred metres away, there are substantial variations variations that directly impact the reliability of crop management, pest control and irrigation decisions.

A single station returns an average. An average that ignores surrounding variability and fails to capture the extreme conditions that most significantly affect the plant’s metabolism. In viticulture, using average data can be as misleading as having no data at all. Today, with LoraWine, all of this has moved from the theoretical world of research into everyday operational practice.

What we found with over half a million data points

Over two years of field trials with the Politecnico di Torino and DISAFA, we deployed a network of IoT sensors across approximately 50 hectares in Langhe, Monferrato and the Turin hills — in collaboration with leading pilot wineries including Cantine e Poderi Oddero, Ettore Germano, Luciano Sandrone and Cantina La Maranzana.

The overarching conclusion: every vineyard has at least one zone with structural anomalous behaviour — persistent and seasonally predictable. Not an exception — a rule. The important thing is to know it.

The full results were presented at the LoraWine launch event, held in March at the University of Turin campus in Alba. The complete report is available here: read the article

The real challenge: not "whether" to deploy sensors, but "how"

Knowing that a distributed sensor network captures far more useful information than a single weather station was, of course, the easy part. The real challenge was finding an effective methodology for deploying sensors across an entire viticultural territory, at a contained cost.

• Connectivity in areas of complex orography — Hillside slopes, valley floors, plots on opposing exposures: signal coverage in these contexts is unpredictable. The most critical micro-climatic conditions are often found in locations where conventional technological solutions fail, yet these are precisely the most important positions to monitor.
• Optimal density: neither too many nor too few — There is no universal rule applicable to every vineyard. Optimal density depends on the extent, terrain morphology, soil type, natural features and specific monitoring objectives. Too few sensors and you fall back into the “problem of averages”. Too many and the cost exceeds the added value generated.
• Data that don’t talk to each other — Sensors from different manufacturers, different protocols, different platforms. The typical result is an information fragmentation that makes it impossible to have an integrated view of the vineyard. Each device becomes an island, and correlating data from different zones remains impossible without a level of integration that almost no commercial solution currently provides natively.

LoraWine is not an adaptation of generic solutions to an agricultural context. It is a service designed specifically for the wine sector and its particular requirements. The two-year trial, beyond the technical-scientific validation of our methodologies with the Politecnico di Torino and DISAFA, allowed us above all to work in close contact with some of the most influential players in the sector. This enabled us to understand and deepen our knowledge of the context in which growers operate, and to define better ways of proceeding together.

• Real connectivity, not theoretical — The network is continuously validated on the ground and covers every necessary zone in the vineyard and cellar, not just where it is easiest to install.
• Density calibrated to the plot — The network configuration is not standardised but sized according to the specific characteristics of the vineyard: morphology, soil variability, monitoring objectives. The research results have enabled us to build a tool to define this with precision.
• Integrated platform — Data from different sensors, positioned at different points in the vineyard, flow into a single system that makes otherwise fragmented information comparable. A complete overview of each plot — treatment coverage status, optimal intervention windows, overnight temperature swings during critical phases. Everything accessible in one place.