Below the list of projects cofunded by the María de Maeztu program (selected via internal calls, in this link the first one launched at the beginning of the program, and in this link the second one, launched in September 2016).

In addition, the program supports:

The detail of the internal procedures for the distribution of funds associated to the program can be found here

Back

Knowledge Extraction for Retail

Knowledge Extraction for Retail

Knowledge Extraction for Retail

The Ubiquitous Computing Applications Laboratory (UbiCA Lab) and the company Keonn technologies work since 2013 in using robots and RFID to inventory retail stores. The scope of this research has been broadened along time to make it more general, more ambitious, and more challenging, given the fact that other groups at the Department have sound complementary expertise to target these challenges. An initial current focus is to explore the integration of the expertise at the Artificial Intelligence Group in algorithms for robot navigation and mapping

 

The overall objective is to develop systems using robots and/or drones equipped with RGB-D cameras, laser scanners, RFID readers, and all other sensors needed to capture all information that is necessary to recreate a digital replica of a retail store, including layout, fixtures, and products, on which to develop business use cases such as inventorying, planogram inspection, in-store picking, or customer assistance. To advance towards the above objective, we need to face several unsolved research challenges, including:

 
  • Unassisted exploration: ideally, the robots and/or drones should be able to start operating by themselves in a new store with no previous information and without the need of a map or any previous calibration.

 
  • Cooperative exploration: several robots and/or drones should be able to work together, sharing the necessary information to explore the store.

 
  • Robust exploration: new algorithms and new sensors (e.g. RFID), previously not used in robotics, should be explored to make the exploration task robust in all environments.

 
  • Sensor fusion: RFID, RGB-D cameras, lasers, ultrasound, IMUs, RTLS, beacons, and other sensors should be used together to obtain the most accurate model of the store, including layouts and planograms, product location and orientation, pricing and publicity labels, or safety related information.

 
  • Multimodal exploration: robots and drones should work together, the robots providing navigation assistance and a power source for the drones, which in turn extend the robot’s reach into more difficult areas.

 

To know more:

 

 

Principal researchers

Rafael Pous

Researchers

Víctor Casamayor (PhD Student)