Events, timelines, states, teams combined with Industrial performance and scheduling.
Brief start about Gantt charts
When designing Dashboards or Reports, Time is often overlooked and used only in Time Series, but there is a lot more to it when employed in Gantt Charts.
Gantt charts are really interesting to use as they emphasize time and event insights.
In this article, we are going to explore its history, set up the notions, and apply a Gantt chart step by step with some tips at the end.
Rollback in History
Figure 1. Adamiecki’s Harmonogram.
Speaking about project scheduling, the article “The Harmonogram of Karol Adamiecki” cites that [1] The first known tool of this type was developed in 1896 by Karol Adamiecki, who called it a harmonogram.
Yet it was published in 1931 and in Polish so it is less known today compared to Gantt Chart who according to Thomas Telford [2] the chart is named after Henry Gantt (1861–1919), who designed his chart around the years 1910–1915 for systematic, routine operations.
It was even applied during World War 1 as mentioned in “The Gantt chart: A working tool of management.” [3] at the instigation of General William Crozier.
Figure 2. Gantt chart example.
Now let’s dust the history and explore Gantt charts step by step.
Building a Gantt Chart
Dashboards and Reports might present Gantt Charts to improve performance by analyzing machine states, production scheduling, and teams planning. Let's bring up the bricks together and come up with a Gantt Chart example.
Notions
First things first! We should define a common language when talking about a Gantt chart to be sure we have a solid base.
1 Tags: Tags represent an entity in which events may occur in a timeline.
2 Time: Dimension commonly visualized as a line.
2.1 Start: starting time of an event.
2.2 End: ending time of an event.
3 Events: An event is an action or occurrence delimited by our start and end times.
Now, let’s map our notions to a finished Gantt chart example.
Figure 3. Gantt chart example explained with notions.
We visualize our concepts using pixels. It is half of our visualization process. The other half is preparing data.
Data
Today, tables are still widely used as logical representations of our stored data. Let’s have a look then at how a simple and ideal table should contain our data before being visualized.
Figure 4. Gantt chart data example used for Figure 3.
Preparing data is a whole topic aside from this one, but empathizing on this aspect strongly improves our overall visualization process.
Understanding step by step
In order to visualize our Gantt chart, based on our previous sections, the steps we follow are:
Understanding what makes a Gantt chart a Gantt chart ( events, tags, and time ). Tags speaking abstract, are entities that may perform actions or welcome occurrences measurable by time. In our example, our tags are machines. They perform actions ( to work, to stop, and to pause ) over a period of time.
Feel the data. For now, we can stick to our table representations of data and try to associate it with the chart. We can use simple data samples and check if the table is coherent with the essence ( see the notions ) of our Gantt chart.
Check if our Gantt chart sticks with the insight we seek to present to our audience. Adapt our chart and use the potential of the chart to its very end.
For our last point for instance we might transform our Gantt chart as below:
Figure 5. Advanced Gantt chart with the
same configuration as in the previous examples.
Let's interpret our example.
Reading it out
In figure 5, if we look at machine 1, we can see that it was working between 08:09 and 09:09 on a Friday. Now, let’s suppose that our goal or question is to present our machine performance to our audience.
We can clearly see that our visualization is not enough to fulfill our goal because there are gaps in our timeline: we don’t know the state of our machines 1 and 2 in the given timeline. So whether our data is missing or we didn’t choose the right question.
Now let’s look at some Gantt chart design recommendations.
Quick tips
Figure 6. Gantt chart example with multiple tags and events.
#1 Event and tag frequency.
We should avoid saturating our chart with voluminous tags and events otherwise as we might see, it becomes hard to read.
#2 Event colors
Taking the same example, we should try to apply different colors when we got distinct event “types”, for example “working and paused “.
#3 Event time delimitation
We should avoid events missing end or start dates otherwise, we might display non-relevant visualization as in our example in Figure 6.
Conclusion
Gantt Charts, are awesome charts thanks to their intuitive reading and help us gain insight into the state of our entities throughout a time period. Some insight examples we acquire with these charts are micro-stop detection of machines that are harder to detect on aggregated Key Performance Indicators (KPI), overseeing downtime durations, analyzing team performances.
Reference
[1] Marsh, E.R. (1975). The Harmonogram of Karol Adamiecki. Academy of Management Journal, 18(2), pp.358–364.
[2] Scirp.org. (2013). P. W. G. Morris, “The Management of Projects,” Thomas Telford, London, 1994. doi10.1680/mop.16934 - References - Scientific Research Publishing. [online] Available at: https://www.scirp.org/(S(i43dyn45teexjx455qlt3d2q))/reference/ReferencesPapers.aspx?ReferenceID=853729 [Accessed 24 Feb. 2022].
[3] Clark, W., 1922. The Gantt chart: A working tool of management. Ronald Press Company.