For three decades, the Line Chart has dominated industrial equipment monitoring. Temperature, pressure, vibration, current. Whenever time-series data needs visualization, the default is a curve. But take motor winding temperature as an example. What operators actually care about is not the average. They want to know whether the temperature is climbing steadily or spiking suddenly, and whether the fluctuation range is widening or narrowing.
Line Charts cannot answer these questions. The answers are often buried in dense curves. Engineers may need to zoom, pan, and compare multiple time ranges before they can tell whether a meaningful fluctuation is present.
Starting with TDengine Historian v1.0.19, the Visualization Panel officially supports the Candlestick Chart. This is not about piling on chart types. The goal is to make fluctuation visible as a concrete pattern, rather than leaving it as something engineers have to infer from dense curves.
Data density is changing how we look at data
The logic of a Line Chart is straightforward: take one value at each point in time and connect them into a line. For the past thirty years, this has been the default choice for industrial monitoring. When data was collected once per minute or less, each point represented the state of that entire minute. String them together and you get an intuitive trend.
But that picture has changed. High-frequency data collection is now standard on industrial sites. Sampling every second, or even every millisecond, means the granularity that once covered “one minute” can now be broken into tens of thousands of pieces. Data density has jumped by orders of magnitude. The analytical habits have not fully caught up.
The problem is that a Line Chart is inherently a “one value per moment” chart. It cannot express what happened at different times within a single window.
When data density is high enough, users want finer-grained information. Is the temperature climbing steadily or spiking and then falling back? Is the pressure fluctuation range widening or narrowing? What is the interval and pattern of intermittent shocks in the vibration signal?
These questions share a common trait. They are not about a single point in time or the overall direction of a trend. They are about “the structure and rhythm of change within a time window.”
A Line Chart compresses hundreds or thousands of data points inside each window into a single visual element: one dot or one line segment. The information structure within the window gets flattened in the process.
A Candlestick Chart uses one visual element to preserve four key positions within a window: the start, the end, the high, and the low. It does not replace the Line Chart. It supplements the dimension that the Line Chart leaves out: the structure of fluctuation inside the window.
The industrial semantics of the Candlestick Chart
The Candlestick Chart has been used in financial markets for centuries. Each time window encodes four key values: Open (the value at the start of the window), High (the maximum value within the window), Low (the minimum value within the window), and Close (the value at the end of the window).
Transplanting the OHLC logic into an industrial context yields the following semantic mapping:
| Financial Term | Industrial Term | Example: Motor Winding Temperature (1-hour window) |
| Open | Window Start Value | Temperature reading at the start of the hour |
| High | Window Maximum | The instantaneous peak temperature within the hour |
| Low | Window Minimum | The lowest temperature within the hour |
| Close | Window End Value | Temperature reading at the end of the hour |
The visual structure of a candlestick consists of two parts. The body, a thicker bar, represents the range from the start value to the end value. The wicks, thin lines extending above and below, represent the extreme values reached within the window. When the time span is long and candlesticks become too dense, users can switch to SPVE Bars. This style is less visually distinctive than candlesticks, but it supports higher information density and is better suited to scanning long-term trends.
This structure maps to three patterns of fluctuation behavior in industrial scenarios.
Pattern 1: Long body, short wicks. The window experienced a significant net change, but the extremes did not deviate far from the start and end points. The parameter shifted steadily without violent oscillation. This is a “trend-type window”: the equipment is drifting consistently in one direction.
Pattern 2: Short body, long wicks. The window saw little net change, but the value briefly touched an extreme far from the starting point. A “shock-and-recovery” event occurred. This is a “transient event window”: the parameter deviated momentarily before the system or operator pulled it back.
Pattern 3: Large and expanding body and wicks (cross-window pattern). Unlike Patterns 1 and 2, which are visible in a single candlestick, degradation is recognized by comparing consecutive windows. When body and wick ranges are expanding window after window, system stability is deteriorating. This is a “degradation trend.”
On a Line Chart, identifying these three patterns requires manual zooming, panning, and visual estimation.
On a Candlestick Chart, you can distinguish them at a glance.
Industrial application scenarios
Line Charts still matter. But relying on them alone can hide critical signals. Here are four scenarios where the difference is visible.
Motor temperature: distinguishing sustained temperature rise from intermittent shocks
Large motors have winding temperature monitored in real time. A Line Chart might show that “the temperature occasionally hit 70°C over the past week,” but it cannot tell you whether the cause is sustained overload or sporadic shocks.
What the Line Chart misses: Within a one-hour window, the temperature sits at 50°C, spikes to 72°C for three minutes, then recovers to 52°C. On a zoomed-out Line Chart, the spike is visually compressed by the surrounding 57 minutes of normal data. An engineer scanning the display is unlikely to notice it.
What the Candlestick Chart reveals: The same window appears as a short body with an extremely long upper wick. The short body indicates that overall temperature did not change significantly. The long upper wick reveals the momentary spike. If this pattern recurs every few hours at regular intervals, you can infer whether the shocks correlate with specific operations.
Decision impact: A long body (sustained temperature rise) points you toward the cooling system. Long wicks (intermittent shocks) point you toward load transients, operating procedures, or lubrication conditions. The two patterns demand entirely different maintenance responses.
Bearing vibration: detecting intermittent shock signals
Rotating equipment vibration sensors read normal values most of the time. But if the bearing cage develops a small fracture, each rolling element passing the fracture point generates a momentary vibration spike.
What the Line Chart misses: On a 30-minute Line Chart, the spike is a barely noticeable bump. It is difficult to tell whether it is an isolated event or the beginning of a repeating pattern without zooming in and comparing intervals across multiple windows manually.
What the Candlestick Chart reveals: Using a 30-minute window, most windows show normal, short candlesticks. But every several windows, a candlestick appears with an extremely short body and an extremely long wick. The normal body means overall vibration has not worsened. The abrupt wick means a transient shock occurred. Regular spacing between these candlesticks aligns closely with the physical model of a rolling element passing a fracture point.
Decision impact: Bearing damage can be detected during the early stage, the “potential failure phase,” before the baseline vibration value rises noticeably. The shock interval, combined with rotational speed and bearing geometry, can be compared against known bearing characteristic frequencies to identify whether the damage is on the inner race, outer race, cage, or rolling elements. This catches the problem at least one maintenance window earlier than waiting for the baseline to lift and trigger an alarm.
Pipeline pressure: distinguishing safety valve actuation from a persistent leak
Pipeline pressure fluctuates normally but occasionally drops rapidly and then recovers. On a traditional Trend Chart, this appears as “a downward spike.”
What the Line Chart misses: Multiple V-shaped spikes may look similar, but their physical causes differ. A safety valve opening and then reseating normally produces a different signature than a slow flange leak. On a zoomed-out Trend Chart, both are just “a dip.”
What the Candlestick Chart reveals: A safety valve actuation appears as a short body with a long lower wick; by the end of the window, pressure has recovered close to the starting point. A persistent leak appears as bodies stepping downward: each window closes lower than it opened, and the cumulative trend points clearly downward. Cross-window coloring can highlight a sequence of continuous declines, making the direction of deterioration immediately visible.
Decision impact: The frequency of long lower wicks quantifies safety valve actuation counts for preventive maintenance scheduling. Cross-window coloring automatically highlights continuous degradation trends, quickly separating normal safety valve operation from suspected leakage.
Batch processes: one batch, one candlestick
In chemical, pharmaceutical, and food and beverage batch production, critical process parameters vary continuously within each batch. Quality departments typically check whether the batch-to-batch mean is consistent, but they overlook differences in within-batch fluctuation.
What the Line Chart misses: Overlaying 50 batch temperature curves on a single trend line produces 50 tangled threads, unreadable. The usual workaround is to extract only the mean or final value per batch and plot a Bar Chart. But this completely discards within-batch fluctuation information. An abnormal batch that runs hot in the first half and cold in the second can have the exact same mean as a normal, steady batch.
What the Candlestick Chart reveals: Each batch becomes a single candlestick. Open is the batch start value, Close is the end value, and High and Low are the extremes reached during the batch process. Fifty batches become fifty candlesticks lined up side by side. Normal batches cluster within a tight range. Abnormal batches stand out clearly. In addition, placing the candlestick chart next to a per-batch sample count chart immediately exposes batches with anomalous sample volumes.
Decision impact: Candlestick height becomes a new process stability indicator. Even if the batch mean stays within specification limits, a pattern of expanding candlestick heights signals deteriorating within-batch controllability. A sample count that deviates significantly from the norm warrants investigating whether the data collection process itself is functioning correctly.
Candlestick Charts were absent from industry not because they were hard, but because the timing was not right
Candlestick Charts have existed in financial markets for centuries. Technically, there is no obstacle to implementing one in industrial software: front-end charting libraries have supported them for years.
Why did no one bring them into the industrial context before? The data environment had not reached the point where they made sense.
Ten years ago, most industrial sites collected data once per minute or less. An hour-long window contained only 60 data points. A Line Chart was perfectly adequate. The Candlestick Chart’s ability to capture fluctuation structure within a window had almost no room to demonstrate its value.
That has now changed. High-frequency data collection is standard. An hour-long window contains thousands, even tens of thousands, of data points. The granularity users care about has shifted from “did the temperature exceed the limit at any point last week” to “how does the temperature fluctuation pattern between 3:00 and 4:00 this afternoon differ from normal.”
The depth of industrial data analysis is also changing. Predictive maintenance, process capability analysis, and root cause tracing all demand more data dimensions. Volatility is no longer just background noise. It has become a diagnostic signal in its own right.
The value of the Candlestick Chart in the industrial domain does not come from it being a new chart type. It comes from the fact that industrial data has finally reached the stage where it needs one. When data density and analytical depth cross a certain threshold, this compressed “four dimensions per window” representation shifts from being a financial-domain specialty tool to a natural choice for the industrial domain.
TDengine Historian introduced the Candlestick Chart at this moment not to differentiate on chart type count, but because industrial users face a real dilemma: data volume is growing quickly, but the visible dimensions of analysis have not grown with it. The Candlestick Chart is an effective path to making the hidden dimension of fluctuation visible again.
The Candlestick Chart has been available in the TDengine Historian Visualization Panel since v1.0.19. For more information, visit idmpdocs.tdengine.com.
