Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS

Safe, Efficient and Integrated Indoor Robotic Fleet for Logistic Applications in Healthcare and Commercial Spaces

Article Category

Article available in the following languages:

The multitasking mobile robots ideal for medical settings

The ENDORSE range of mobile multitasking robots are designed to perform in harmony with humans or other machines, in indoor or outdoor environments.

Digital Economy icon Digital Economy

While robots are already mainstream in warehouses and factory production lines, they are also becoming more common in more ‘unstructured’ environments. “Robots will soon have to co-exist and collaborate with humans and other machines, such as drones, inside and outside,” says Nacim Ramdani, coordinator of the EU-funded ENDORSE project. Focusing on medical settings, ENDORSE wanted to help ensure that multitasking robots could safely navigate this cyber-physical world. The team developed algorithms for human-aware autonomous robot navigation and fleet management systems (FMSs), allowing robots to collaborate towards a common aim, while pursuing individual tasks.

Multitasking ‘infrastructure-less’ mobile robots

ENDORSE wanted to advance robot navigation beyond current lidar sensor approaches with so-called ‘deadlock’ systems, where robots register objects in their path, then stop and wait until the route is clear. “Our robots have pre-installed deadlock resolution algorithms. Using imaging techniques these detect humans and anticipate their movement so the robot can manoeuvre to complete tasks,” adds Ramdani. The team designed a solution deployable without significant infrastructure and investment, minimising installation time while keeping costs down. The ENDORSE FMS is accessed through cloud-based software-as-a-service, avoiding the need to install hardware components on-site, with the solution accessible anywhere with internet connectivity. With an initial target market of medical environments, an e-diagnostic point-of-care kit was developed. This comprised off-the-shelf diagnostic sensors, alongside a proprietary ECG sensor, coupled with a telecommunications (network gateway device) and commercial-grade software (developed by the project) to connect sensor measurements to patients’ electronic health records. The team also built different mounting gear which could be fitted to the robots, allowing them to perform various mechanical tasks, such as transporting a trolley.

Testing times

The 3D robot simulator Gazebo was used to simulate the movement of multiple robots in a range of indoor environments, to assess both the efficiency of the FMS, as well as specific autonomous navigation modules, including the deadlock resolution and robot localisation methods. The FMS also underwent extensive simulations and experiments, using both metaheuristic strategies and optimisation algorithms, to find the best set-ups and parameters for allocating subtasks to individual robots. Multiple simulations were also run to transmit sensor-captured (dummy) patient data to cloud-based electronic health records, to validate the interaction and integration of these medical modules. The team also tested actual robots in a realistic set-up. “A two-robot fleet successfully performed tasks assigned by the system, with the robots autonomously navigating an office space shared with humans,” explains Ramdani. “These experiments revealed new challenges. For example, glass-wall partitions, often present in offices, can refract sensor signals, degrading performance of the robots.”

A wide range of applications

Since the 1980s, robots have provided surgical assistance thanks to their precision and tireless working. But advances, such as computer vision, artificial intelligence and big data, have vastly extended robot capability. Interest is growing in caring and rehabilitative robots to help patients with illnesses, cognitive impairments and disabilities. In 2020, the International Federation of Robots estimated the medical robot market would be worth USD 9.1 billion by 2022. “Robots could reduce the burden on healthcare professionals, freeing up time for specialist duties, improving overall quality of care while reducing costs,” notes Ramdani. Partners are now assessing how to introduce the ENDORSE solution into the healthcare market – specifically elderly homes, clinical centres and hospitals. As a Marie Skłodowska-Curie Actions project, knowledge transfer through interdisciplinary staff exchanges was at the heart of ENDORSE. This helped spawn a sister project RESPECT, now working to address any security and privacy issues identified during ENDORSE.

Keywords

ENDORSE, robot, fleet management, autonomous, navigation, safe, algorithms, artificial intelligence, medical, hospital, health

Discover other articles in the same domain of application