IEEE Robotics & Automation Magazine - March 2021 - 80

section, the disinfection scenario was composed of two
sequential treatments. The first was composed of three
PoIs, shown in Figure 8(b) and represented by an " a "
marked by orange dots. The latter had only two PoIs,
which are indicated by a " b " marked by blue dots. The
increasing number represents the order in which the robot
traveled to the stations. The executed trajectories are
reported with orange and blue dashed lines, respectively,
and the disinfection stations are indicated by large orange
and blue dots. From Figure 8(b), it is possible to observe
that the robot reached all the PoIs and, at the end of each
procedure, returned to the charging station.
The metrics computed for the disinfection scenario are
presented in Table 5. For the scenario, the estimated working time was roughly 2 h. The disinfection of the corridor
was carried out at night,
generally from 2 a.m. to
4 a.m. for safety reasons,
The presence of the
and it was performed
35 times in two months.
robot had significant
During the procedure,
the robot traveled 85 m
psychological benefits for
for the first phase and 30 m
for the second one. The
patients forced to stay in
disinfection scenario
was easier to be booted:
the treatment center and
only one health-care
operator was involved.
away from their families.
He or she had only to
select the desired disinfection procedure and
turn on the UV-C light. At the end of each phase, the robot
returned to the charging station.
Introducing the robot into highly biohazardous environments had multiple advantages in both scenarios.
First, in the logistics scenario, the robot reduced the
number of movements required by medical personnel
among different environments. This led to a reduction in
the biological risk to which the operators were subjected.
In addition, the presence of the robot had significant psychological benefits for patients forced to stay in the treatment center and away from their families. The robot did
not aim to replace the warmth and moral support of
health-care professionals but enhanced their role and
increased their benefits.
Table 5. The results of the disinfection scenario.
Scenario 2: Disinfection
Number of performed procedures

Mean time disinfecting

125.6 ± 3.4 min

Mean traveled distance

116.4 ± 7.2 m

Number of health-care operators

Usage indexes of the proposed robotic platform are reported as the
mean value and the standard deviation.

IEEE ROBOTICS & AUTOMATION MAGAZINE

MARCH 2021

Despite this, using the robot had uncovered weaknesses. The platform runs on battery power and needs to be
periodically recharged. Medical personnel must be adequately trained to properly work with the robot, which is
not easy during a global pandemic. Finally, it is better to
define the map before using the navigation algorithms.
So, during the first installation, expert robotics users
have to go through all relevant areas with the robot to
record the map and manage the navigation and obstacleavoidance algorithms.
Conclusions
In this article, a two-month experiment studying the use of
the UCBM COVID robotic system was described. The
robot was used to assist health-care operators during their
daily activities, in particular, for logistics and disinfection
procedures. To accomplish such tasks, the robot had
autonomous navigation capability. The robot was
equipped with the Gmapping SLAM algorithm to provide
robustness against a dynamically changing environment
and facilitate the ease of implementation. A list of the
parameters used to implement the navigation capabilities
was provided as a guideline for further developments of
autonomous robots for similar scenarios.
Preliminary tests conducted in laboratory settings
were carried out to assess the robot's autonomous navigation capabilities. The good results obtained in a controlled environment enabled the robot to access the
UCBM COVID treatment center. To address two scenarios with the same platform, a modular system was developed: the robot could be equipped with a box for
delivering materials and with a UV-C lamp for disinfection. Furthermore, two web GUIs were developed so that
users could easily and intuitively select commands for
each of the two scenarios.
The introduction of the robotic system facilitated the
parallelization of material transport inside the ward, and
the proven effectiveness and ease of use led operators to
exploit the robot throughout the work day. Users did not
have to be expert to interact with the robot, thanks to the
GUIs. During the night shift, the robot was exploited for
disinfection purposes. The absence of health-care workers
in the corridors made it possible to carry out this activity
under safe conditions. The automation of the procedure
enabled treating the entire area of interest without human
intervention.
Future efforts will aim at increasing safety during the
disinfection procedure and improving the interaction
modalities between the health-care operators and the
robot. In particular, software and hardware improvements will be implemented to automatically detect people entering an area that is being disinfected and turn
off the UV-C lamp. Moreover, voice commands and
control are being developed to improve the interaction
modalities and communication between robot and
health-care personnel.

IEEE Robotics & Automation Magazine - March 2021

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2021