Industrial Activity monitoring for the RACE portal.
COVID-19 has had a severe impact on the globally integrated automotive industry. Large scale manufacturing interruptions across Europe is an example. Several shutdowns during the “lockdown” months of March, April and May generated big losses (4.024.036 motor vehicles from January to September 2020, according with the European Automobile Manufacturers Association, ACEA).
For the Rapid Action on Coronavirus and EO initiative, we have validated the parking occupancy indicator as a proxy of car industrial activity.
Furthermore, we have compared the model performance obtained with Medium Resolution data (Sentinel 2) versus High Resolution data (SPOT), observing a very good agreement in the tendency.
By using Sentinel 2 open (free) data, the cost of the monitoring is dramatically decreased, while the performance is still high. This is a great achievement for scalability.
One more year the ESA Φ-WEEK brings us the latest state of Earth Observation.
Our vision about DL4EO shared in the Space & AI Conference and the AI & Big Data Congress.
We will talk about the current state, barriers and future directions of Deep Learning for EO.