A conceptual framework for developing dashboards for big mobility data

Figures & data

Figure 1. Comparison of visual analytics platforms and dashboards in the three key dimensions: input data, analytical complexity, and user interaction for output display.

Figure 2. Conceptual framework for big mobility data dashboard design and development.

Table 1. Mobility data types and suggested visualization elements in a mobility dashboard.

Data types	Visualization element
Categorical data: movement attributes	Pie chart, line graphs, symbols, and bar or radar charts that show statistical summaries and comparisons (W. Chen et al., 2015)
Time series data	Line graphs, heatmaps (e.g. calendar views), radial diagrams, or histograms that represent periods or cycles, while sequential aspects can be conveyed through calendars or timetables (S. Chen et al., 2019; Kveladze & Agerholm, 2020; Wang et al., 2020)
Network: transfer links (S. Chen et al., 2018; Kveladze et al., 2019), origin-destination (OD) flows (Koylu & Guo, 2017)	Graph charts, flow maps
Moving object	Map with moving points, map with geometric trajectories
Spatial-binned data: big trajectory data; density maps (Kveladze et al., 2019)	Map with quadtree-based module

Figure 3. A dashboard for a vehicle insurance company to monitor driver behaviors.

Figure 4. A dashboard for a fleet management company to monitor the activity of trucks. a) the view for fleet operators; b) the view for high-level managers.

Figure 5. A dashboard for a hospital to evaluate resource-allocation model results.

Figure 6. Interface of Toronto Transit dashboard (https://totransit.ca/).

Figure 7. Interface of Euro-Argo project dashboard (https://fleetmonitoring.euro-argo.eu/dashboard).

Table 2. Main input data sources and their generalized trajectory models for Dashboard 1.

	Vehicle trajectory	Crash event	IMN
O	Drivers as insurance policy
P	GPS waypoint	Location	Place locations and ordered location pairs
T	Timestamp	Timestamp	Month
A	Driving events	Crash or not	Visitation frequency and origin-destination flows

Table 3. Data selection processes for Dashboard 1.

Selection function	Visualization element on the dashboard	Input and selected output
$O\times P\times T\to P$	Map	Given driver ID (O), binned GPS waypoints (P), timestamp (T), the binned GPS waypoints of a driver within a user-chosen time period. December is selected in this example.
$O\times T\times A\to A$	Graph chart	Given driver ID (O), timestamp (T), and individual mobility network (IMN) as O-D flows (A), a driver’s IMN in a specified month is selected.
$T\times A\to \hspace{0.17em}T\times A$	Bar chart	Given timestamp (T) and crash event (A), the monthly numbers of crashes are selected.
$O\times T\times A\to A$	Text	Given driver ID (O), timestamp (T), and crash risk probability (A), a driver’s crash risk probability in a specified month is selected.

Table 4. Main input data sources and their generalized trajectory models for Dashboard 2.

	Vehicle trajectory
O	Vehicle IDs of a fleet
P	GPS waypoint
T	Timestamp
A	Driving events, engine statuses, and modeled indices

Table 5. Data selection processes for Dashboard 2.

Selection function	Visualization element on the dashboard	Input and selected output
$O\times P\times T\text{rarr}P$	Map of Fig. 4a	Given binned GPS waypoints (P), timestamp (T), the binned GPS waypoints of all vehicles (O) on a specified date is selected.
$T\times A\text{rarr}T\times A$	Line chart of Fig. 4a	Given a range of timestamp (T) and number of daily trips (A), the number of daily trips during a year (for example) is selected.
$A\text{rarr}A$	Text box of Fig. 4b	Select the statistics (number of trips) of the current periods (i.e. day, week, month) and the corresponding previous periods.

Table 6. Main input data sources and their generalized trajectory models for Dashboard 3.

	Population/patient density	Current facility	Proposed facility
O	Region ID	Facility ID	Facility ID
P	Region location	Facility location	Facility location
T	-	-	-
A	Density	Distances to assigned region	Distances to assigned region

Table 7. Inferred input data and their generalized trajectory models for Dashboard 4.

	Vehicle trajectory
O	Vehicle ID
P	GPS waypoint
T	Timestamp
A	Heading, direction, route, and seconds since the last update

Table 8. Inferred input data and their generalized trajectory models for Dashboard 5.

	Float trajectory
O	Float ID
P	GPS waypoint, depth
T	Timestamp
A	Direction, salinity, temperature, country, year of deployment, etc.

Figure 8. Interface of jobs in the Netherlands dashboard (https://dashboards.cbs.nl/v1/commutingNL/).

Table 9. Inferred input data and their generalized trajectory models for Dashboard 6.

	Commuter flow
O	Flow ID
P	Origin location, destination location
T	-
A	Jobs (as the number of commuters), origin municipality name, destination municipality name, region names, etc.

Chen, W., Guo, F., & Wang, F. Y. (2015). A survey of traffic data visualization. IEEE Transactions on Intelligent Transportation Systems, 16(6), 2970–2984. https://doi.org/10.1109/TITS.2015.2436897

 Web of Science ®Google Scholar

Chen, S., Andrienko, G., Andrienko, N., Doulkeridis, C., & Koumparos, A. (2019). Contextualized analysis of movement events. International Workshop on Visual Analytics, 49–53. https://doi.org/10.2312/eurova.20191124

 Google Scholar

Kveladze, I., & Agerholm, N. (2020). GeoVisual analytics for understanding the distribution of speeding patterns on arterial roads: Assessing the traffic safety of vulnerable road users. Journal of Location Based Services, 14(3), 201–230. https://doi.org/10.1080/17489725.2020.1823497

 Web of Science ®Google Scholar

Wang, Q., Lu, M., & Li, Q. (2020). Interactive, multiscale urban-traffic pattern exploration leveraging massive GPS trajectories. Sensors, 20(4), 1–16. https://doi.org/10.3390/s20041084

 Web of Science ®Google Scholar

Chen, S., Wang, Z., Liang, J., & Yuan, X. (2018). Uncertainty-aware visual analytics for exploring human behaviors from heterogeneous spatial temporal data. Journal of Visual Languages & Computing, 48, 187–198. https://doi.org/10.1016/j.jvlc.2018.06.007

 Google Scholar

Kveladze, I., Kraak, M. J., & van Elzakker, C. P. J. M. (2019). Cartographic design and the space–time cube. The Cartographic Journal, 56(1), 73–90. https://doi.org/10.1080/00087041.2018.1495898

 Web of Science ®Google Scholar

Koylu, C., & Guo, D. (2017). Design and evaluation of line symbolizations for origin-destination flow maps. Information Visualization, 16(4), 309–331. https://doi.org/10.1177/1473871616681375

 Web of Science ®Google Scholar

Data availability statement

Data used for Dashboard 1, 2, and 3 are not available due to both legal and commercial restrictions. Dashboard 4, 5, 6 are from third-party authors who are not affiliated with this manuscript. Data availability of Dashboard 4, 5, and 6 is subject to the creators of the dashboards.