“Data is the new currency” is a catchphrase heard throughout many industries. According to the World Economic Forum, by 2025, the amount of data generated each day is expected to reach 483 exabytes globally.
In our daily lives, data is generated every time someone Googles for information, asks Alexa for something or accesses their phone or smartwatch. For the industrial automation sector, Global Market Insights projects industrial sensor shipments to grow past 3 billion units by 2026. The growing implementation of industrial sensors, along with advanced controllers and data acquisition platforms, is enabling real-time data collection at extraordinary rates, creating exponential growth.
However, collecting the data is not enough, it needs to be visualized in context as useful information for quick and easy interpretation. Good information allows users to establish meaning, find patterns and anomalies and draw conclusions. Today, with the large amount of data received from sensors, controllers, IIoT devices and other systems, advanced data visualization becomes paramount.
The human mind processes visual data better than text or other types of data, and the human brain also remembers stories and visuals better than text-based information, so providing information in these formats is critical.
The big picture
Several user interface (UI) and data visualization design trends are beginning to impact industrial automation, including concepts such as using real-life pictures, aesthetic minimalism and sophisticated typography. Effective UIs are needed for local human-machine interfaces (HMIs), site-wide supervisory control and data acquisition (SCADA) systems, and overall equipment effectiveness (OEE) displays — and show results of advanced analytics and predictive maintenance. These UIs need to capture the attention of the operator, be intuitive and allow quick interpretation of the data so action can be taken.
A Microsoft study shows the average human’s attention span in the year 2000 was 12 seconds, and by the year 2013 this had declined to 8 seconds. Therefore, designers must incorporate pre-attentive attributes as they create the data visualization so operators can notice screen information without using undue conscious effort.
Pre-attentive attributes are defined as attributes our brains process in milliseconds, such as color, form, spatial positioning and movement.
Color. Color can either enhance or distract from the data visualization message. The ISA-101 standard for HMI design recommends using color sparingly, and mostly for representation of anomalies or abnormal states. Situational awareness methodology — making the user aware of the context when designing screens and dashboards — also advocates for using color and shapes sparingly to ensure the operator’s attention is drawn to actionable data.
When using color for data visualization, experts also advocate using color consistently and effectively, with an understanding of the cultural implications of the palette selected. The meaning of color varies depending on culture or color biases. For example, in Western culture, red is often used to signify alarms or danger, while in China red represents good luck and success. Therefore, when designing data visualization screen, the effect of color needs to be considered in context with the message being conveyed. Color should be used purposefully to either draw attention to and highlight a portion of data, or to distinguish among various data points.
Brightness is related to color, and dark data dashboards — black backgrounds with pastel or neon color palettes — are growing in popularity as inspired by futuristic films such as Star Wars, and major brands such as Apple. Dark data dashboards prolong display device battery life and improve visual ergonomics by reducing eye strain, and automatically adjusted brightness based on current light conditions facilitates use in dark environments. Because of these and other advantages, dark data dashboards are now used more than those with light color themes (see Figure 1).
Form. Form applies to attributes distinguishing data on a screen. For example, dynamically increasing the size or prominence of an object or screen area signifies importance and quickly captures operator attention.
Spatial Positioning. Spatial positioning is the space used to represent data, and it may include the use of two- and three-dimensional representation. Most data visualization is depicted in two-dimensional space. However, geographic information system and three-dimensional representation is effective for representing geographically dispersed assets. With smart devices and sensors being deployed everywhere, spatial positioning is taking on greater importance for achieving data visualization operational efficiency.
Movement. Movement has always been a common way to capture operator attention. However, caution must be used to avoid annoying or distracting movement, flickering and pop-ups. Consider how web advertisers attempt to capture our attention with dynamic displays, and how many of these ads become annoying and prompt users to engage ad-blocking software.
Properly implemented, UIs can benefit from animated buttons on a data dashboard as a powerful technique for indicating to operators when action is required. This is especially important for space-constrained mobile devices.
Trends in data visualization
The use of proactive, continuous intelligence for automatically processing real-time and historical data also can play a role by allowing operators to process increased data volumes. By detecting anomalies and producing effective UI indications, alarms and events, data visualization becomes an effective way of alerting operators of potential issues. Here are some additional data visualization trends.
Smart devices. Smart devices are everywhere, collecting data all the time and contributing to the proliferation of data. Yet the value of data diminishes as it ages, so data visualization must occur closer to the data origination, while providing visualization to the enterprise level. Furthermore, connection across distributed data sets allows organizations to identify valuable connections, relationships and insights among these disparate devices.
According to IDC, by 2024, “70% of enterprises will have integrated their edge-generated data with cloud-based enterprise systems to allow for real-time actioning based on IoT analytics.” The incorporation of smart devices to provide real-time data analysis, along with visualization closer to where the data is generated, while incorporating these data sets into corporate architecture, provides decision makers with the information they need, where and when they need it.
Mobile devices have become indispensable in our lives and will likewise play an integral data visualization role in enterprise digitization. Mobile devices possess limited graphical real estate, requiring data visualization on these devices to be clear, simple, compact and concise. As attention spans wane, data visualization on mobile devices must provide awareness while simplifying interactive features.
Storytelling. Data visualization using traditional HMI/SCADA and dashboards show operators what is happening, but not why. Data storytelling is a method for communicating the why with a compelling narrative. Data storytelling puts information in context, reducing the chance of incorrect interpretation (see Figure 2).
Gartner forecasts indicate data storytelling will be the primary way users consume analytical data by 2025. Data storytelling leverages data, narrative and visuals tailored for the operator to effectively communicate actionable insights.
In other words, data visualization is representing data graphically, while data storytelling is putting meaning behind the data. This provides a narrative consolidating a complex amount of data in a simple, concise manner, resulting in a story explaining the how and why behind the data.
Real-time visualization. Visualizing data in real time is nothing new for industrial automation applications. However, given the increasing velocity of data collection and the growing number of deployed smart devices and sensors involved, the need has expanded. Real-time visualization must collect, aggregate, correlate and provide analysis to accelerate the decision-making process. Real-time analytics will become paramount to help users understand what is happening. As devices become smarter, the data they communicate will contain more information (see Figure 3).
Curating effective data communication
According to Forbes, 35 billion smart devices will be online by 2021, growing to 75 billion by 2025. Data visualization is becoming more important as the volume of data is increasing while attention spans decrease. Data effectiveness will increase as data visualization becomes more intuitive and data storytelling principals are applied.
Data visualization has reached new heights with the emergence of cognitive frameworks, which are defined as a set of elements from psychosocial theories on how humans process data. Dashboards for HMIs, SCADA, OEE, advanced analytics and predictive maintenance must evolve to allow the visualization of large amounts of data, while providing the best techniques for visual communication. Effective data visualization is intuitive, offers easy visual recognition and enables quick understanding of the data being presented.
Data storytelling will become more common in industrial dashboards for communicating real-time data as well as analytics. Machine learning techniques for data visualization will continue to ascend into traditional industrial markets for optimizing actionable decision making.
Data fuels the power of intelligent automation and strategic decision making. There is a beauty in data visualization when it simplifies curated data and conveys contextual information in a manner appealing to users. The right data visualization software coupled with intelligent configuration will deliver these and other benefits to provide improved operations throughout the enterprise.