Instrumentation & Measurement Magazine 25-1 - 50

Field
Measurements
and Analysis
Fig. 3. Software Architecture (Task Distribution).
necessary tools to use Wi-Fi-TCP communications. The ZEBRevo
GeoSLAM 3D LiDAR SLAM system provides its own
local portable wireless network and ROSCore node. The two
separate networks can handle high bandwidth streaming
from camera feeds and the 3D LiDAR platform. A TP-Link
AC1900 wireless router was used to augment the GeoSLAM
3D LiDAR SLAM system local network. Statistic IPs were
assigned to all the computing platforms in the developed
system to communicate effectively over the ROS communication
bridge.
Software Modules
The software architecture of the system was based on the
distributed ROS environment. The implemented software
comprises two main interactive software packages: the data
acquisition and processing software (DAPS) and the remotecontrol
software (RCS). Fig. 3 shows the tasks' distribution of
the implemented software.
Data Acquisition and Processing Software
(DAPS)
The DAPS combines the ROS nodes, which run on the embedded
platforms that control the sensors and prepare the
sensory data to be transmitted to the RCS. The DAPS package
is also designed to perform pose estimation based on the
logged sensory data, offloading any extensive optimization
calculations to the remote server. The modules running within
the DAPS package are shown in Fig. 3. For each sensor, one or
more ROS nodes was developed in C++ under Linux to configure
and control the sensor, then encapsulate each data frame
into the proper ROS message and then transmit it to the RCS.
This distributed processing framework guaranteed real-time
performance.
Remote Controlled Software (RCS)
As shown in Fig. 3, the RCS runs on the server and is managed
by the Qt framework, known for its real-time efficiency
and suitability for embedded systems with a graphical user interface
(GUI). The general software tasks implemented by the
RCS are summarized as follows:
◗ It provides an easy-to-use GUI.
◗ It initiates, control, and monitor the system modules.
◗ Logging received ROS messages into a ROS bag file.
◗ It provides real-time 3D visualization.
50
Sensors and Network
Setup
The system was tested and
measured at Carleton University
in Ottawa, Ontario.
Fig. 4 shows the system
with body-mounted sensors.
Fig. 5a shows the
entire test trajectory overlaid on the Carleton University campus
map; the green portion is outdoor, while the blue is indoor.
It included multi-floor stair walks to increase experiment
and dataset versatility. Fig. 6 shows the UWB anchors
mounted at pre-defined locations in the Minto Building lab
area where the experiment started.
Upon exiting the lab, the test navigated the Minto Building,
including four floors of stairs. The testing path then continued
outside, exiting the building from the southwest side and reentering
from the northeast side, returning to the lab area. The
test trajectory was designed with several loop closures to reduce
errors and drifts in SLAM [16]. Thus, the GeoSLAM 3D
Instrumentation & Measurement Magazine 25-1

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 25-1
