HomePage

Meet Bill

TDP 2025
Qualification Video 2025 (a)
Qualification Video 2025 (b)
GitLab 2025

TDP 2018
Qualification Video 2018

TDP 2017
Qualification Video 2017

TDP 2016
Qualification Video 2016

The ACSO - Center of Computer Architecture, Intelligent Systems and Robotics at the State University of Bahia (UNEB) has been participating in RoboCup with the ACSO BahiaRT team since 2009 in leagues such as 2D Soccer Simulation, Mixed Reality, 3D Soccer Simulation, and @home, among others.

The service robot proposal of ACSO BahiaRT for the RoboCup@Home league is called Bill (Bot Intelligent Large capacity Low cost), codename Estranho. It was born in 2014 as a result of research projects in assistive robotics. Its main functions are communicating with humans through natural language processing, recognizing objects and faces, and navigating through unknown environments. ACSO BahiaRT with Bill got 13th place in 2015 RoboCup@Home and 21st place in 2016. Specifically, in RoboCup@Home Brazil, got second place in 2015, 2016, third place in 2017 and seventh place in 2024.

Over the years Bill has been improved to better meet the @home challenges. Now we presents the third generation of Bill, named Bill Estranho, and its main improvements. This year, a great part of the ACSO BahiaRT effort was dedicated to building Bill Estranho. This work involved re-engineering the hardware using new components, redefining the architecture, and changing the operating system to ROS2. As a consequence, all the other functionalities, i.e. voice recognition, face and object recognition, navigation, and manipulation needed to be updated

For better accessibility in human-robot communication, we bring the contribution of recognizing the Bazilian Sign Language (LIBRAS). Another contribution regarding human-robot communication, our team brings the emotion detection by speech. This research aims to generate better answers according to the emotions detected in the human speech. Concerning object detection, we are currently working on the integration of Yolo and Tesseract and still have some results that indicates an improvement in detecting objects that have labels.

Bill is part of the RoboCup @Home initiative which aims to promote research on assistive robot technology and the development of fully autonomous robots capable of interacting with a domestic environment. He has several skills to help him with his purpose, such as communicating with humans through natural language processing, recognizing faces, navigating through the environment, among many others. His name was carefully chosen to represent his main characteristics: Bot, Intelligent, Large Capacity, Low Cost.

The best thing about Bill is that he is not expensive. He was fully built inside our lab with low cost components, making him not only powerful, but also affordable. He can definitely do a lot with little.

Evolution of architecture

UNEB CNPq ACSO BahiaRT BILL