News

The UAH hosts the 21st edition of the International Workshop of Physical Agents where artificial intelligence will be the protagonist

On November 23 ends this international online forum focused on research on artificial intelligence (AI) techniques in control and interaction applications with physical agents,such as mobile robots, vehicles, manipulators, etc.

On Thursday 19, at 9 a.m., the opening session was attended by Javier de la Mata,vice-chancellor of Research and Transfer of the UAH, and Luis M. Bergasa,principal organizer and professor of the Department of Electronics.

In total there are 5 sessions, which can be followed in this link where researchers from all over the world meet to promote and popularize the scientific foundations of physical agents, for example mobile robots, vehicles, manipulators, biobots, etc. today, controlled with Artificial Intelligence (AI). AI is expected to revolutionize our daily lives in the near future and the unstoppable 'robotic revolution' will open up a myriad of challenges for its application over many and varied physical agents.

The International Workshop of Physical Agents is sponsored by the Network of Physical Agents, Springer and Robocity2030 and addresses a wide range of research challenges in the fields of multi-agent systems, software agents, human-robot interaction, mobile robots, social robots, cooperating robots, autonomous vehicles, machine and deep learning, perception, localization, mapping techniques and autonomous navigation systems.

As in previous editions, the organizing committee supports presentations with live or video demonstrations and exhibitions and will feature presentations by leading subject matter experts. During today's closing Friday, awards will also be held for the best works presented during the meeting.

In recent years, the deployment of the autonomous car has been gradual, constant and has become normal. News of vehicles driving themselves through a neighborhood or testing on the highways has ceased to amaze. From wonder to curiosity. The prodigy has become an encompassing phenomenon, not always understood, but less arcane and rather technical.

In the middle of 2020 there are already many trials in real scenarios. It has even been tried to create commercial services - with a touch of experimentation but with all the guarantees - based on the transport of people in autonomous vehicles. At this point, the data collection is already large, the technology has been perfected, both the sensors and the artificial vision, the information processing or the algorithm training. Apparently the machines are practically ready. But at the technological level - the legal and social levels follow their own times, more slowly - there remains a detail.

There are many voices that point to the need to have 5G for the autonomous car to reach the highest level. This is what is known as level 5, which allows full autonomy. Here the driver's attention, even his presence, is no longer necessary. And is that the vehicles of this category can be manufactured even without steering wheel or brake or accelerator pedals.

In this article we can find some important points such as:

  • Real-time decision making: the importance of latency

    • What is necessary for a level 5 autonomous car

  • Large volumes of data

    • Information management in the autonomous car

  • V2X: Vehicle to everything

    • The problem of obsolete car fleets

    • Smart cities, autonomous car umbrella

    • First experiences with the autonomous car

  • Remote drivers

    • The debate on the regulation of the autonomous car

If you want to read the news, click on the following link: https://blogthinkbig.com/coche-autonomo-nivel-5-5g

During the month of November 2020 our colleague Carlos Huélamo win the First Prize 🏆 in the "2020 Awards - Campus of International Excellence: Bioenergy & Smart Cites" in the Master's thesis category, funded by the Universidad de Alcalá and Universidad Rey Juan Carlos.

The project, titled "Predictive Techniques for Scene Understanding by using Deep Learning", was developed in the RobeSafe research group (Department of Electronics, Escuela Politécnica Superior UAH) under the outstanding supervision of Prof. Luis Miguel Bergasa.

The scope of the project is to propose an accurate and Real-Time Multi-Object Tracking (MOT) software architecture in the context of Intelligent Transportation Systems (ITS) by means of sensor fusion between LiDAR and Deep Learning based Visual Object Tracking. Additional tools involved in the project were ROS (Robot Operating System), Docker, Inc, the CARLA (CAR Learning to Act) autonomous driving simulator and our real-world autonomous electric vehicle.

 

To finish we leave you a phrase that Carlos cited regarding the award:
"Going together is starting, staying together is progress, working together is success"

During the month of November 2020, our colleague Eduardo Romera obtained the First Prize 🏆 in the "2020 Awards - Campus of International Excellence: Bioenergy & Smart Cites" in the Doctoral Thesis category, funded by the University of Universidad de Alcalá and Universidad Rey Juan Carlos.

The project, entitled "Semantic understanding of 3D scenes for autonomous vehicles through deep learning", was developed at the  Escuela Politécnica Superior UAH under the outstanding supervision of Prof. Luis Miguel Bergasa.

 

In September 2019 Luis Miguel Bergasa gave a lecture at the AI and Robotics Workshop organized by RoboCity2030 in Madrid titled "Robocity2030: Madrid Robotics Digital Innovation Hub".

In this WS the poster presented by our group entitled "Designing a Drive-by-Wire System for an Autonomous Electric Car" was awarded the prize for the best WS poster.

 

Location

Polytechnic School

University of Alcalá

Av. Jesuitas s/n.

Alcalá de Henares, 28871

Madrid, Spain

Smart Elderly Car

Sarbot Team