This project originated in the Robotics Research Centre of Nanyang Technology University (NTU) located in Singapore. Both faculty members and PhD students are involved with many aspects of robotics in areas of healthcare, environment and construction. With increased digitalization and government funding, the construction industry in particular shows promise of improved productivity and would benefit from the application of mobile robots.
Professor Chen and his team of six focused their energy in this field to bring Quicabot to life. The Building Quality Inspection and Assessment Robot, or Quicabot for short, is a quality assessment robot used in construction sites. This robot based on Jackal UGV, uses various sensors to inspect construction locations for defects, such as cracks, flat walls and hollow tiles. The robot compares the information from the inspection with industry standards, and makes the assessment to confirm if the area passes inspection.
To gather this information, Quicabot is equipped with a number of sensors, including:
- RGB Camera for visual inspection
- 2D LiDAR sensors to detect the evenness of walls
- Inclination meter to measure the slope of the bathroom/balcony and drainage
- Infrared Camera to check for wall hollowness
The team integrated the sensors themselves and installed another layer of architecture for assessment on top of the Robot Operating System (ROS), which is used for navigation. When a robot is deployed, it is usually working with a human assessor. However, navigation can be autonomous or tele-operated. The robot can be pointed from one location to another, can roam autonomously, or can be made to follow the assessor.
When considering mobile platforms to use, the team looked for something lightweight and portable. A more robust platform was necessary for commercialization – that’s where Jackal UGV comes in. Professor Chen and his team chose Clearpath’s Jackal UGV platform firstly because it is compact and easier to move around large areas, which is perfect for large buildings and sites. In addition, the ruggedness of the Jackal was ideal for the semi-finished buildings and construction environments. The Jackal UGV was also chosen due to its ROS integration and open architecture.
Collecting Quicabot’s data helps assessors obtain more detailed information to achieve 100% inspection, and is collected at a speed 2x faster than a human. According to Professor Chen, Quicabot reduces the amount of labour needed by approximately 50%. He also estimates a 30% reduction in the cost of assessing and surveying construction sites, based on market surveys and robot performance.
The team worked on the project from 2016-2017, then created the spin-off company Transforma Robotics. They are currently commercializing construction related robotics and continue to work on R&D.