Background: Evaluation of Data Visualization methods is a major challenge within the software and other industries. With complex data and requirements, often organizations require effective visualization methods which impact business decisions and convince stakeholders. This is a similar challenge in the development of a software demonstrator for the innovative safety radar system at ABB Jokab Safety whose aim is to improve the detection reliability using multiple radar sensors and requires an effective visualization method which will satisfy all the requirements.

Objectives: The main objective of this study is to explore different data visualization methods involved in illustrating the raw data and with the help of developers, and other team members feedback with reference to existing literature and filter them with respect to the system functionalities. Establish evaluation criteria with relevant metrics to perform analytic evaluations on the visualization methods to determine an effective method.

Methods: A Case Study which includes a multivocal literature review, is conducted at ABB Jokab Safety. Initially, to gather information on the subject, both formal and grey literature are explored and documented to filter our appropriate data visualization methods for this system. A task-based evaluation using semi-structured interview is conducted on 14 participants to determine an effective visualization method followed by statistical analysis to derive proper validation of the findings. The Response time, Ease of understanding, Confidence and Accuracy of the visualization methods are evaluated with feedback from the participants.

Results: The Multivocal literature review filtered 16 primary articles which encouraged the use of 4 data visualization methods used in the safety radar system with distinct functionalities. A coordinate transformation engine to combine the data sets was also developed for the safety radar system contributing to the overall improvement of detection reliability. The evaluation including both quantitative and qualitative results validate each other’s findings through statistical tests like Kruskal-Wallis and Bonferroni post hoc followed by narrative analysis resulting a heat temperature plot to be more effective in the visualization of the radar data from multiple sensors.

Conclusions: The results from this research provide insight into how data visualization evaluation can be implemented for real-time industrial problems and furnish validation process to determine an effective data visualization method. This study helps object detection using similar radar technologies visualize their data in an effective way and provides a scientific approach for evaluating similar data visualization problems.