Engineering: leader della Digital Transformation

Engineering Innovation in

Smart Energy & Utilities


Snam: Advanced Analytics for Gas transportation

Our new approach increases the efficiency of "unaccounted-for gas" (UFG), enabling actions for the improvement of the whole system.

Approach & Solution



Within the scope of the SUMMER programme, launched to provide better access to the commercial and plant information of each redelivery point, Snam needed to find a new approach to manage the issue of Unaccounted-for Gas (UFG).

UFG is the quantity of gas remaining from the balance between the gas injected and the gas delivered to end users, net of line-pack variations and of the consumption connected with the operational management of the network.

It has been highlighted that it has a highly variable trend over time and, hence, Snam has carried out several analyses of UFG determinants throughout the years, in order to address actions to reduce such variability.

Continuous discussions and the skills introduced in the measurement sector, which also relate to the chemical / physical analysis of the quality of gas, had identified several variables that influenced the thresholds of the meters.

It was therefore necessary to adopt a more scientific approach, using Advanced Analytics techniques to solve the problem.


Engineering, parallel to the SUMMER programme started by Snam, owned both process and tech skills needed and hence began an initial prototyping activity for approximately one year, in order to collect useful data for the subsequent project phases. Once the analyses were completed, we identified six main use cases for Snam on which to focus our analyses. We adopted a methodological approach for the construction of the data mining model (Crisp-DM) that progressively allowed to identify correlations between UFG and sub-threshold measurements.

We carried out continuous discussions with Snam on the results of the analyses, which we carried out with the help of machine learning activities, allowed to better understand these phenomena. Hence, we began an industrialisation plan of the models implemented and defined a roadmap to complete the planned cases in 2022.

The architecture defined is based on the Microsoft technology stack: Data Lake Gen2 for data storage, Synapse for data structures, and Azure Machine Learning for the front-end Advanced Analytics.

The user interface was designed as consistent with the choices made for SUMMER, identifying ways of representation appropriate to the market best practices and allowing users to create standard reports independently and autonomously.


Effective interaction with ARERA authority

Reduced Unaccounted for Gas

Improved quality of data analysis

Project value

Process performance

Enabling Technologies

AI & Advanced Analytics

Project Team

Data & Analytics
Energy & Utilities
Engineering Interactive