Executive Summary

DT Report in 1 Minute

Automation and Artificial Intelligence

It is not about robots like androids … Key AI tech has no physicality: automatic pilots of airplanes, AI systems in surgical interventions… Now we transfer AI to manufacturing (autonomous vehicles) and transport (autonomous trains).

It is in autonomous vehicles that drive by themselves; in a public lighting system that optimizes energy; in an algorithm processing stock market transaction; and in expert systems personalizing medical treatments. Or our word processor helping us write.

Robots and AI are not directly related. A robot is an entity that executes orders from an intelligent” algorithm. AI is “cognitive abilities supported by technology”.

Artificial Intelligence or human intelligence?”. The goal is improving human intelligence with AI, not AI replacing human intelligence. Automating repetitive tasks, minimizing errors, improving accuracy, AI increases speed of execution. Goal is not to eliminate human intervention, but its inconveniences.

Synergy of human intelligence + AI = “Augmented Intelligence”. But it’s really “Augmented Human Intelligence“, boosted by AI, taking advantage of its speed and certainty processing data to make decisions based entirely on data.

Adjuvant Technologies

Big Data has become a fundamental tool in the context of IoT (or IoE) and Industry 4.0: a population of robots working autonomously and generating data on a large scale. Incorporating AI, they make the right decisions for themselves…

Big Data is based on Data Lakes in the cloud, where we store all data. When the robot generates data, we move it to the cloud to be processed. This slows down real-time answers and robots may be rendered inefficient.

New method: use the same data to make the same decisions, without physically transporting and storing it anywhere else. Many of these data are not necessary after decision is made, so they can be removed.

Edge Computing’s answer: the data a robot needs is processed at the robot, without moving it elsewhere. This quickens response time, simplifies communications and gives each robot its own autonomous intelligence.

Drones now move in groups without bumping into each other: a central entity not needed for coordination. A step in upgrading the individual intelligence of robots.

Use case: Healthcare

New era of personalized attention based on Artificial Intelligence.

Thousands of images for medical tests + diagnoses used to train algorithms to identify tumors or malfunctions.
Algorithms compare symptoms and evolution of patients, analyzing them over time.
Deep Learning
Machine Learning algorithms are not invasive: they try and learn without danger for patients.
Digital Twins
Health insurance’s key investment: clients not falling ill.
Predictive Analytics