Objectives

General

  • Improve the team's deep learning capabilities
  • Extend INESC TEC networking capabilities through the partners’ contacts and scientific meetings with the key players in the market.
  • Bring a fresh and new research perspective to some of the already existing problems in today’s and future field robotics applications such as: search and rescue, underwater deep-sea mining, submerse mine mapping, forestry, aquaculture, agricultural robotics, advanced driving systems, aerial surveillance, aerial mapping, etc.

Operational

  • Reinforcement of scientific and technological human potential of INESC TEC staff.
  • Promote the cross-fertilization, i.e., interaction between people not only from different institutions but also with expertise in different topics.
  • Enable a stronger networking between INESC TEC research staff and world top researchers.
  • Provide an opportunity to bring together groups of national and international research staff and world top experts, discussing advanced topics related to deep-learning.
  • Provide opportunities for developing scientific and personal relations (due to summer/winter school duration).
  • Promote early stage researchers careers.
  • Provide INESC TEC research staff with appropriate skills in order to allow them to work at different technology readiness levels (between 3 and 8).
    Potentiate the creation of new joint research initiatives.
  • Provide opportunities for establishing contact between young researchers and world top experts as well as faster scientific exchange mechanisms.
  • Enable the training of research staff in specific topics.
  • Ensure the involvement of all relevant stakeholders.
  • Ensure the dissemination of project results to scientific, industry and investment communities, as well as to the public. It aims also at the exploitation of project results.
  • Promote the exploitation of project results.

Functional

Autonomous Underwater Manipulation

Increase INESC TEC’s capabilities to improve and develop new approaches for autonomous underwater manipulation allowing its application in INESC TEC unmaned underwater vehicles (UUVs) for field robotics missions underwater (e.g: mining, oil and gas, security…).

Navigation and autonomy with high-level scene understanding

Increase INESC TEC’s capabilities for robust and resilient long-term navigation and autonomy through robot perception, capable of enabling numerous field applications using autonomous/semi-autonomous robots.

Semantic mapping in ground field robotics application

Increase INESC TEC's competences in ground field scene reconstruction and ground field context awareness exploiting semantic information extracted via deep learning techniques. Semantic mapping is indeed the result of synergic interactions between deep learning semantic scene interpretation and mapping. To overall objective is to show how semantics can provide the proper priors for a better and more accurate reconstruction of the environment and, at the same time, the reconstruction itself can be used to boost semantic interpretation, e.g., via deep learning on 3D structures such as graphs.

Aerial Robotics - Target Tracking and Control

To develop and deploy novel multi-robot cooperative target tracking methods for a team of aerial vehicles addressing coastal survey, search and rescue applications.
To develop cooperation strategies for heterogeneous teams of robots involving aerial and water surface vehicles.
Applications include autonomous landing on floating platforms for self-recharging, cooperative detection of people in dangerous situations (e.g., drowning...)