To touch or not to touch......

[dwalker 0]

I 'm working on automating a small blender in our plant. I have about 6 inputs and 6 outputs ranging from start/stop pushbuttons to a motor starter and stack light. I'm going to use a small touch screen to allow the operator to setup blend times so I guess I want to know...... How do other people approach using a touch screen as your start/stop buttons? There will be a hardwired E-stop that will disconnect all output power and motor starters, but is there a conflict by using a touch cell as a start or stop button? I would think, and if I'm interpreting NFPA79 correctly, that if the there is logic to ensure that the touch screen is communicating with the PLC and that if in case of a loss of communication the PLC would stop operation then I should be ok. How does everyone else handle touch screen vs. hardwired buttons? Thanks. Dave

[ssommers 20]

I prefer having at least one hardwired set of normal start & stop buttons to ensure that the machine can run in auto and stop regardless of the HMI. My reason... E-stops can be messy since you're simply removing power from everything. Having a normal stop button allows for an orderly shutdown with or without the HMI. I've had an HMI screen backlight die in the middle of running and my operator was able to remember where the touch buttons were in order to shutdown the machine, but I don't want to do that again. Machines can usually live without changing setpoints for a day or so when an HMI dies or you can change the setpoints via the programming software. Living without a main start & stop button is much more difficult. I know that start & stop takes up 2 inputs. That can be a pain when you're looking at a tiny application like this, but make room at least for a hardwired normal stop button if a sequential shutdown is necessary.

[Ken Moore 23]

I think every situation is unique. If you can tolerate an E-stop without quality issues, then I would go with out hard stop/start station, because you could always use the E-stop if needed. If having a sequential stop is vital to product quality, then the hard stop is justified. Since you mentioned a blender, I would guess, a sequential shut down isn't required, so a hard stop would not be needed. My $1.298's worth.

[TConnolly 13]

Me too. For the same reasons ssommers gave.

[paulengr 44]

NFPA 79 only requires that you shut down on loss of controls in the case of wireless controls such as crane pendants. HMI's are pretty much exempted. Generally speaking it depends on your degree of trust in the HMI. When you do your risk assessment, whether you need an extra layer of how to shut it down should pop out. This is when you do the what if component X fails scenarios. In manufacturing lines, it is common to have the extra hardwired input. But in process controls, it is frequently the case that although there might be multiple HMI's, there might in fact be no practical way short of turning off the PLC, throwing disconnects, or using some sort of E-Stop to actually shut down the process. Frequently operators have access to multiple workstations so this isn't as bad as it sounds. NFPA 79 does not REQUIRE an E-Stop. They tell you how you must implement it if you have one but don't require it. They do require some way to stop the whole system. But that's a totally different issue. Furthermore, it requires that you stop all ACTUATORS. It doesn't mean cut power to everything. That's a common misinterpretation. You can leave all your indicator lights and inputs live. However, you've got to do your homework first. Crank out that risk assessment. So far every time I do a risk assessment, depending on the safeguards you provide, you will frequently find that an E-Stop is not necessary. Furthermore, I really dislike them for a lot of reasons. They tend to get abused as a function for a lot of other things. For an E-Stop to work properly, it must emergency stop...ie, if the equipment is utterly destroyed in the process, that's OK. If there's something that would move by gravity, it must have some sort of brake to prevent that from happening. An example of a real E-Stop operation is when you scram a nuclear reactor. This triggers the safety system to slam ALL the control rods into the reactor core all at the same time, and flood the area with cooling water. And, this is the part that 99.999% of the world misses for E-Stops, you must, train, train, train, and retrain. You've got to train this like it's a military operation and your operators work off reflexes. Because in an emergency stop situation, the normal automatic human reaction is either freeze or flight. The last thing your operator will typically do is hit an E-Stop because that's not the natural reaction. So you've got to train until it becomes an automatic reflex. So far, I've never actually seen anywhere except military teams and perhaps some mine rescue teams actually train to that level of emergency procedures. Otherwise, your E-Stops totally lose their functional purpose. It is far better to design guarding, lockouts, interlocks, etc., which totally eliminate the need for E-Stops. At this point the only reason that I place them in the equipment is because some managers go totally bat blank if you even suggest eliminating them completely. So when they ask for E-Stops, then I insist that we do them by the book. When I start asking for operators to be trained and certified and set up a huge paperwork trail and do retraining/certification constantly, they suddenly see the light and decide that maybe E-Stops were a bad idea after all. In addition, if someone does trigger an E-Stop, this means an accident actually happened which became an emergency situation. Because that's the nature of an emergency stopping situation. So you must do an accident investigation and file all the correct paperwork. By nature, the reset switch should be pretty much under lock and key. So where's your paperwork? This is another reason that most managers suddenly decide that E-Stops are a really bad idea...suddenly they are threatened with another layer of needless bureacracy. Edited 15 Jul 2008 by paulengr

[Clay B. 1]

I tend to add Start /Stops for another reason. To save wear and tear on the HMI. These buttons are used the most and when an operator is having a bad day, take the most abuse.

[ssommers 20]

This is where I disagree... guarding, lockouts & interlocks LIMIT the need for e-stop points, but I don't think they can every COMPLETELY eliminate the need for them in every situation. A risk assessment is still KEY to determining places where an e-stop actuator makes the best sense.