Touchscreen Requirements for Human Machine Interface in Industrial Automation
As touchscreens become commonplace in HMIs, the requirements for a touchscreen controller have to evolve to specifically target this market.
The Human Machine Interface (HMI) in industrial automation is quickly transitioning to touchscreens. The touchscreen requirements on a factory floor, however, are unique and must be solved for a good (read “safe”) operator experience with increased productivity and throughput. Three of these requirements are waterproofing, noise immunity, and advanced touch capabilities like glove touch and/or proximity sensing.
As more countries look to manufacturing as a key economic growth driver, the industrial automation market is forecasted to reach more than $200 billion by 2015 from $159 billion in 2012 (IMS Research).
This market has gone through a transition with regard to user interfaces in the last decade. Instead of switches and levers, interaction between humans and the automation equipment, today called the HMI, is via 3.5” – 10” touchscreens.
According to Frost & Sullivan, in factories of the future, many of these controls will be enabled through handheld devices wirelessly talking to the machine. With the rapid adoption of touch as a de facto user interface, a touchscreen on these handheld devices is almost a necessity. And, given performance requirements, this touchscreen will have to use projected capacitance technology instead of resistive.
Factories are starting to focus more on the quality of these user interfaces. HMI today is not a background control interface but serves as an identity of the machine or process. A poorly designed HMI can lead to false inputs and introduce delays resulting in errors in processes, damage to equipment or product, or even operator injury. At the worst, a poorly implemented touchscreen leads to operator frustration. In contrast, a robust HMI leads to increased productivity and throughput for greater profitability.
Factories pose unique challenges to HMI and touchscreen design. Three of the key challenges are:
- Waterproofing: Prevent accidental touches due to water, tracking a finger through a water droplet, or tracking a wet finger.
- Noise Immunity: Ability to provide a seamless touch experience and prevent false touches in the presence of extreme noise spikes.
- Advanced Touch: Ability to work with gloves and detect a finger in proximity to the screen.
Waterproofing is an often overlooked, yet extremely important requirement for a robust and reliable user-interface. Many manufacturing environments have high levels of humidity, and operators may be required to operate the touchscreen with water on their fingers or on the screen. The touchscreen has to work smoothly and not trigger false touches due to this.
There are international standards that cover waterproofing for touchscreens in detail. The International Electrotechnical Commission (IEC) standard IEC-60529, for example, contains definitions for Ingress Protection (IP) ratings. One of the highest ratings a product can have is IP-67, meaning it can handle a significant amount of dust (dust rating of 6) and can be immersed in water up to 1m (water rating of 7) without damage. For most industrial applications, this should be considered a must-have.
From a touchscreen controller point of view, “waterproofing” (this term applies to all forms of liquid or conductive particles on the screen) can be further broken down into two requirements: water rejection and wet finger tracking.
- Water Rejection:
- Prevents the touchscreen from accidentally responding to the presence of liquid.
- Enables seamless operation when liquid is wiped off the screen.
- Wet Finger Tracking:
- Accurate touch sensing when liquid is present on the screen. This could be in the form of a thin layer or film of liquid due to humidity, a spill, or multiple droplets.
- Touching the screen with a sweaty or greasy finger.
Projected capacitance works by sensing the change in capacitance when a conductor steals charge from a grid of metal (typically Indium Tin Oxide, or ITO) wires separated from each other, which act as sensors when current is passed through them. These wires are arranged as Tx (where current is passed) and Rx (where it is received), and there is capacitance formed between the Tx and Rx wires.
There are two implementations of projected capacitance. Self-Capacitance sensing (Self-Cap) detects changes in charge at the rows and columns (X + Y) of the sensor grid. As charge change in a given row could be attributed to multiple columns, self-cap is used for single-touch applications.
Mutual-Capacitance sensing (Mutual-Cap), on the other hand, detects changes in charge at every intersection of the grid (X + Y). Hence, multiple touches could be precisely sensed. See Figure 1.
A finger touch appears differently in self- and mutual-cap sensing modes. In self-cap, a touch is registered by an increase in current once the charge is transferred to ground, whereas in mutual-cap a touch is detected as a reduction in the overall mutual-capacitance between the two sensors in the intersection.
Water acts as a conductive medium that strengthens the fringe field between adjacent sensors and increases capacitance. This may cause the touchscreen to register water as a light finger touch in self-cap mode. This can be addressed by sensing an exact replica in the adjacent sensor, which effectively eliminates fringe field coupling between adjacent sensors.
However, self-cap does not enable multi-touch.
Water behaves the same way in a mutual-cap sensor grid, but is sensed as an increased charge, which is in opposite polarity to the effects of a finger touch. This can sometimes lead to the sensor registering a false-finger touch when water is removed from the screen.
A combination of self and mutual capacitive sensing, such as is implemented in Cypress’ (San Jose, CA) TrueTouch controllers, provides a robust waterproofing solution. It is also important to be able to switch the Tx and Rx lines to get an accurate profile of the water droplet.
When the screen is covered with a film of water or a big water droplet, the effect could be similar to a large object like a thumb or a palm (depending on the size of water droplet/film). Special algorithms are needed to precisely determine the position of this water body and track a moving finger.
There are essentially two noise sources for a touchscreen:
- Direct-Coupled Noise: This is noise generated by adjacent machinery, HVAC, and electronic ballasts from CFL lights – all present in manufacturing plants – that get coupled onto the human body and injected into the system by the finger touch.
- Common-Mode Noise: This is the noise generated from within the touchscreen device (e.g., power supply, bad quality charger) and discharged via the human finger to ground. Noise contains both broad-band and narrow-band tones, often with high amplitudes. We have seen common-mode noise spanning frequencies up to 500 kHz and ranging in amplitudes up to 40 Vpp. See Figure 2.
In either case, the user sees false touches; either as false touch coordinates being reported or an overdrive of the touch sensors (a touch appearing as a long streak along the Rx sensor). This could lead to incorrect commands on the assembly line and cause delays. In many cases, the noise spike saturates the receiving capacitor, thereby losing any touch signal that may be recorded in that intersection and impacting the overall touch experience. A good signal-to-noise ratio (SNR) is one of the requirements for a touchscreen controller to combat various forms of noise.
There are several ways to combat noise.
- Higher Tx Voltage: One of the most effective ways to increase SNR is to increase the signal voltage. This is a simple but effective way to improve SNR. Some of the Cypress Semiconductor touchscreen controllers provide a built-in 10V Tx capability that increases SNR and avoids additional BOM cost increases.
- Frequency Hopping: The Rx channel can dynamically change its frequency to avoid noisy bands and their harmonics. Frequency hopping has to be active during noisy conditions, and special algorithms need to be built into a touchscreen controller to dynamically hop from a noisy frequency. Besides the methods listed above, there are several other noise suppression techniques available. Some of these newer methods effectively prevent channel saturation, while recovering the signal using windowing via an on-chip DSP.
- Advanced Touch:For mobile phones, conductive gloves are available in the market to detect touch. This is not an effective solution on the factory floor as the operator may be required to wear special gloves to operate other machinery. It is also inconvenient to have the operator take gloves off for touchscreen operation.
To a host CPU, a glove touch is no different from a weak finger touch. Hence, a touchscreen controller could increase the sensitivity and have a low finger touch threshold to register a touch. However, this may create problems like:
- A hovering finger may be detected as touch, which may not be the intended user behavior.
- Common-Mode noise may trigger false touches.
- Performance will vary depending upon the thickness of the glove.
Besides glove touch, a touchscreen controller may be required to sense an approaching finger when it is as far as 25-30 mm away from the touchscreen. This could trigger an event like turning on the LCD for optimal user experience.
Advanced touch features can be enabled using different sensing methods, special algorithms, and touchscreen tuning, or a combination thereof.
As touchscreens become commonplace in HMIs, the requirements for a touchscreen controller have to evolve to specifically target this market. Industrial users expect their touchscreens to work in the presence of various forms of noise and conductive material like water, gloves, etc. A touchscreen design that takes care of these requirements ensures a great user experience and improves worker productivity, thereby increasing overall factory throughput.