AI Case Studies

We assessed a data-driven model to establish the real value of 30+ production plants for a top Italian player in the energy industry.

The result was increased valuations by a range of 60%-100% across the portfolio, leading to higher profits during sale.

Starting from the asset register: a granular list of all elements, including the more complex items, we were able to provide a reliable estimate of the life span of the components.

By combining proprietary data with statistical data, literature and expert opinion, we could depreciate the assets, based on Objective Useful Life.

An asset manager and equity research company needed to monitor ESG performance and risks relating to the listed companies in its portfolio.

We developed a custom project based on a proprietary ESG framework and the daily collection of news data from different web sources to alert clients to specific ESG risks according to a sentiment score, thus considering the companies’ reputation.

Our customer could monitor ESG risks and corporate reputational critical issues on an intuitive dashboard and a quarterly detailed report.

A leading international automotive player wanted to understand how to manage the allocation of spare parts sales budgets to its 1000+ dealers.

We developed a Decision Support System to estimate the risk for dealers of not reaching the desired budget goals (based on historical data and remaining stock) and provide meaningful insights for the optimal allocation of budgets to each dealer, in an efficiency-centered perspective.

A major Italian banking institution commissioned us an IDM custom project to optimize its document management flow for the consumer credit division and the onboarding of new customers.

By integrating our IDM technology in the existing infrastructure, we allowed the bank to handle 90,000 files a year, totaling more than 2.5M pages from 33 different file types (IDs, insurance forms, contracts etc.).

This way the entire process was made more efficient, from the document upload to the automatic compilation of the relevant fields, getting accurate and immediately actionable data, as well as saving more than €460,000 per year in money.

A top player in the consumer electronics industry needed to minimize the risk related to third party distributors, monitoring news and reviews about them, in order to support risk managers in identifying critical issues that involved specific partners and points of sale.

Our solution 3rdEye and its proprietary algorithm was able to process large volumes of alternative data about each store to identify risk signals on a daily basis, carrying on an accurate sentiment analysis of customers and providing a score for every reseller, highlighting critical issues and specific opportunities.

Our client could rely on a daily ranking of the risk associated with each retailer and shop selling its products, along with timely alerts about critical events detected in real time.

Our customer needed to improve risk prevention in their gas extraction plant for different types of technical Gasses, like Hydrogen and Oxygen.

Available data was a state of the art description of current anomalies more likely to lead to an accident.

Through data modeling we developed a tool that automatically assesses the risk according to failure patterns, in order to prevent critical events.

A major international player in the energy world needed to understand how to optimize streetlight maintenance.

The probability of malfunctioning assets depends on environmental factors (wind, temperature range, distance from the sea) and characteristics of their materials.

We developed a decision support system (DSS) based on an AI algorithm in the form of an interactive interface. By entering geographic areas and lighting-related parameters, our customer can identify those light points that are most worth to be replaced.

A multi utility company needed to prioritize maintenance for the medium voltage network in Milan. The electrical distribution counts up to 6,000 switching stations, connected to each other by so-called branches.

We developed a model to analyze the branches fault history, considering their specific characteristics (types and ages of the materials, number of joints, etc. ) and the availability of alternative branches nearby, to ensure service continuity in case of failure.

This way we highlighted those stations where the probability and the severity of fault risk were higher, in order to prioritize interventions accordingly.

A railway network managing company needed a risk analysis for the circulation of a new hydrogen train on a non-electrified line, in a highly urbanized section.

To understand all the risks of such trains on that railway, we had to consider all possible scenarios that could lead to catastrophic events, combining available data with many more unknown variables in our risk analysis model.

A large pharmaceutical company asked us to analyze their drug encapsulation process, in order to reduce rework times and production waste.

We have revolutionized their approach to data collection, teaching them how to correctly value all signals, instead of mistakenly taking the average.

Distinguishing between anomalous and normal signals, we promptly traced the anomaly (a misalignment of the shaft) and predicted where a production discontinuity was likely to occur.

A major railway transport player needed a new, more efficient weighting systems for trains, to estimate transport costs and ensure safe infrastructures, such as the bridges supporting their load.

The traditional static weighing system on special rails takes too much time and is expensive in terms of man hours.

We developed a system based on special sensors which can be easily installed on normal tracks, to acquire the most relevant signals and turn them into accurate results.