X

Overview of the industrial innovation projects made by the 5th year engineering students

Students in the final year of their engineering cycle presented their innovative project. The jury noted the teams. Back to the event and focus on the three best projects.

Industrial innovation projects carried out by students are related to their specialization. Beautiful achievements after almost five years of study!

Prototypes for companies

SMS telepathy, an indoor drone… Industrial innovation projects can be very surprising! But they are always part of a professional and concrete dynamic.

In the 5th year, it is a question of valuing your work outdoors and in particular towards companies. At this point, it’s important to increase your visibility.

Hear and Know for example offering solutions in the field of supply chain and logistics. Laboratories are also involved. This is the case of the Da Vinci Research Center. Associations are also solicited, for example, De Vinci Durable.

An industrial innovation project is also the result of the teamwork of 3 to 5 students. Who says work, refers to the use of soft skills, these skills that help manage conflict, find solutions to interpersonal problems.

The showroom’s video

Focus on three industrial innovation projects hailed by the jury

Here is the podium of this year’s projects:

On the 1st place of the podium, a Supply Chain project, IoT and Blockchain

The project aims to secure and optimize the path of a product from a Supply Chain by combining the power of the two technologies IoT (Internet of Things – Internet of Things) and Blockchain (decentralized shared registry). The use of sensors (GPS, temperature…) makes it possible to follow in real time the position of cargo but also to ensure that the conditions of its transport maintain its good quality.

Before a product lands on our shelves, it begins a process of long and tedious step sequence that starts from its creation in the factory and ends with its purchase, it is the Supply Chain.

In the era of globalization, most products of our daily lives are exported. In 9 out of 10 cases, merchandise is transported by sea and it is estimated that the cost of processing and administering commercial documentation accounts for nearly one-fifth of ocean freight costs.

In 2nd position, also an Artificial Intelligence project

Audio Driven Mouth Animation (ADMA) and integrates into the Innovation Course of the De Vinci Innovation Center (DVIC). It involves humanizing Artificial Intelligence (AI) by focusing on the mouth via speech synthesis and the oral articulation associated with it.

The project is inspired by the latest research articles from Google, Nvidia, Pixar and Disney that deal with Deep Learning. To realize this kind of model, it is necessary to have a consequent dataset.

Since there is no free dataset for this task, ADMA provides an automatic method to create it.

After training on this dataset, it is enough to provide him a text so that it generates at the same time a voice and its animation buccal.

3rd project laureate, a project in Machine Learning and Finance

Customer satisfaction is a key indicator of a company’s performance. Any dissatisfied customer may have to undermine the reputation of an organization and turn to competition.

The project aims to build a machine learning tool to assess the customer satisfaction of a large banking company. By identifying better and earlier customers unsatisfied, the bank can be proactive and anticipate the expectations of its customers.

An anonymized database with 369 variables and 76,020 observations was used. At first, after statistically studying the meaning of our variables, data filtering was done to make them consistent. Then, different machine learning algorithms were studied under Python: logistic regression, neural networks and decision trees. After evaluating the efficiency of these by building scores (ROC curve and AUC), we chose to use our decision tree algorithm (Random Forest) to develop our tool.

By entering the customer number in the interface, the tool will calculate the probability that it is unsatisfied and, if possible, propose adjustments to reduce the probability of dissatisfaction.

More about the Educational project at ESILV

Related Post