How-to
Running a PLC-hosted visualization on a remote HMI panel PC
Short answer
The PLC-hosted visualization reaches a panel PC through a panel-side PLC HMI client deployment: install the client for the panel platform, confirm the license, run the remote-client deployment that selects the hosting PLC, verify the network path, and start the client from a CE shell command. Downloading TargetVisualization to the PLC only feeds a monitor connected to the controller; it installs nothing on the panel.
On this page
- What runs where
- Check the network path before software changes
- Steps to put the PLC visualization on the panel PC
- Working the case: a CE 7 panel PC with TwinCAT 3.1 and PLC HMI installed
- Choosing between the client, the browser and moving the program
- First-run symptoms and what each one means
- Questions asked
The visualization already runs on the monitor plugged into the PLC, and the panel PC beside the machine stays blank. The screens live on the PLC, but the panel still needs its own client side before it can show them: the PLC HMI client installed for the panel's platform, the license accepted on the panel, the panel-side deployment that selects the hosting PLC, and a shell command that starts the client. Downloading TargetVisualization to the PLC does none of this; it only feeds a display connected to the controller itself. The step most often skipped is documented under Remote operation of a PLC HMI Client, because the word remote reads as if it does not apply. Work the steps below in order, and the PLC keeps the control program while the panel displays and operates its screens.
What runs where #
Treat the system as two independent computers. The PLC executes the control program and hosts the visualization; the panel PC executes the display client; the Ethernet path carries the traffic between them. Four requirements reach the panel before it can display anything: the client installed for its platform, the license accepted on the panel, the panel-side deployment that selects the remote application, and a verified communication path.
Remote names the relationship between the client on the panel and the application on the PLC. It does not mean the panel runs without local software; the client, its deployment and its license all live on the panel.
| Component | Role in this installation | What you verify |
|---|---|---|
| PLC | Executes the control program and hosts the visualization | TargetVisualization added under Visualization Manager, compiled, activated, downloaded; screens show on a monitor connected to the controller |
| Panel PC | Runs the PLC HMI display client | Client installed, license accepted, deployment done, startup command correct |
| Ethernet path | Carries the client's traffic to the hosting PLC | Link status at both ends and any switch, addressing, route |
Downloading the visualization to the PLC installs nothing on the panel, and the shell command starts a process without selecting any application; the deployment is what binds the client to the PLC's screens. Operating a cabinet PC's own desktop from a panel is a different pattern that needs no client deployment at all, and getting RDP from a Comfort Panel to a cabinet PC covers that path.
Check the network path before software changes #
Confirm link status at each port on the panel, the PLC and any intervening switch. Read the addressing on both devices and confirm a route exists between them: the route created from the PLC with Edit Routes proves the two systems can address each other, and it is transport, not selection — it carries the client's traffic and nothing more. A working engineering laptop connection to the PLC is a different path from the panel's connection, and proving one proves nothing about the other. Fix link and addressing faults before the software steps, because no deployment can compensate for a path that does not exist.
Steps to put the PLC visualization on the panel PC #
- Prove the screens first. Run the visualization locally in the engineering environment against the live PLC and operate a few controls. If that check has not been done, testing an HMI simulation against a live PLC walks through it. Deploying a panel is wasted effort until the screens themselves work.
- Confirm the PLC side. Right-click Visualization Manager, add TargetVisualization, compile, activate and download to the PLC. The download takes longer than before, because it now carries the screens, images and text lists alongside the PLC program; if the transfer itself stalls, why an HMI is detected but the project download fails covers that separate problem. When it completes, a monitor connected to the controller shows the screens. That display is your baseline, not the finished job.
- Read the panel's software inventory. Note the operating system build — Windows Embedded Compact 7 on this class of panel — the installed TwinCAT version, and whether PLC HMI is already present. A CE panel can carry TwinCAT 2 as well as TwinCAT 3.1, and the version decides which package and which documentation you use.
- Install the client package that matches the panel platform. A CE panel takes a CAB package matched to the OS build and the installed TwinCAT and PLC HMI versions; obtain it from the vendor for that platform. The package installs the client and its components. It does not carry your screens or machine commands; the project on the PLC defines those, and the deployment step connects them.
- Confirm the license on the panel. The entitlement belongs to the system that runs the client, so to the panel. Installing files alone does not establish licensed operation; confirm the license is present and accepted before moving on.
- Run the panel-side deployment. Follow the documentation section named Remote operation of a PLC HMI Client, as written for the installed versions. This step sets up the pieces that let the client display the remote visualization, and it is the one the word remote hides. Skipping it leaves a client that launches and shows nothing.
- Point the client at the hosting PLC. The deployment and its remote selection name the PLC application the client displays, and the existing route carries that traffic. The client receives the screens and their variable bindings from the PLC over this path; the panel keeps no copy of the project.
- Set the shell startup on CE. Add a shell command to the panel's CE configuration that starts the PLC HMI executable, including the configuration arguments required for the installed version. The startup documentation sits in the vendor information system under Twincat 3/TFxxxx/TF1xxx/TF1800/PLC HMI, in the CE portion. A command with the right executable but missing arguments launches a process that still has no target, so automatic start-up fails even when a manual start worked.
- Verify end to end. With the machine in a safe state per its documented stop procedure, power the panel down and up so the shell command runs. The client starts, connects to the named PLC and displays its screens; operate one control from the panel and confirm the PLC responds.
Working the case: a CE 7 panel PC with TwinCAT 3.1 and PLC HMI installed #
This is the case behind the question, worked to a decision.
Inputs, as reported with the original problem: the panel PC runs Windows Embedded Compact 7; TwinCAT 3.1 and PLC HMI are installed on it; a shell command is set to start PLC HMI; a route from the PLC to the panel exists, created with Edit Routes; TargetVisualization is added under Visualization Manager and downloaded to the PLC, and the screens run on a monitor connected to the controller; the panel-side deployment under Remote operation of a PLC HMI Client was never performed; the panel license was never confirmed.
The rule in words: the PLC-hosted visualization appears on the panel only when four checkpoints are closed — client installed, license accepted, deployment done, path verified. The shell command counts for none of them; it starts the client and selects nothing.
The arithmetic: installation passes and the path passes, so 2 checkpoints are verified. The deployment is unperformed and the license is unconfirmed, so 2 are open. 2 verified + 2 open = 4 checkpoints, and the panel displays only at 4 of 4; at 2 of 4 it stays blank however many times the client starts.
Result: the panel launched the client at power-up and showed no screens.
Decision: below 4 of 4 the panel cannot display the visualization, so close the open checkpoints — confirm the license, then perform the deployment. On this installation the deployment was the fix: once Remote operation of a PLC HMI Client was carried out, the HMI worked, with the PLC still owning the control program.
Choosing between the client, the browser and moving the program #
The deciding criterion is where the display application executes and which controller owns the process data.
| Approach | Screens appear on | What you configure | Decision |
|---|---|---|---|
| Engineering visualization | In the engineering environment | Nothing deployed; open the screens in the project | Development checks only |
| Local TargetVisualization | A monitor connected to the PLC | Add TargetVisualization, compile, activate, download to the PLC | Your working baseline; the display stays at the controller |
| Remote PLC HMI client | The panel PC | Panel-side deployment, license, remote selection, CE shell startup | The right choice when the PLC keeps control |
| WebVisualization | A browser on the panel | The web visualization service plus a browser on the panel | Choose it only after the panel's browser checks out |
| PLC application on the panel | The panel, executing control there | Remote I/O, or an explicit ADS interface between the two programs | Only when the panel must own the I/O |
Commit to the remote PLC HMI client when the existing PLC must retain control and the panel exists to display and operate that application. A browser design is not an automatic substitute for the installed client; on a CE panel, confirm a compatible browser before selecting that route. Moving the control program onto the panel is an architecture change, not a prerequisite of the client: a panel-side program that sends commands through ADS or another protocol needs an explicitly defined data interface between the two programs.
First-run symptoms and what each one means #
Work these in order; each maps one observation to one cause.
- The client launches and the screens stay empty. The deployment is missing. Repeat step 6 and check that the remote selection names the hosting PLC.
- The client is absent after power-up but starts by hand. The shell command's executable path or its arguments are wrong. Match both against the documentation for the installed version.
- The client will not run at all after a package or build change. The package no longer matches the panel platform or software build; reinstall the package matched to what is installed.
- A monitor connected to the PLC shows nothing either. The fault is on the PLC side, not the panel. Redo step 2.
Questions asked
Does the panel PC need the same download the PLC got?
No. The PLC download carries the visualization data that a monitor connected to the controller displays. The panel PC receives a different installation: the client package for its platform plus the remote-client deployment that selects the hosting PLC. The machine's control program never loads onto the panel, and the panel does not take over I/O. Two systems, two installations, one network path between them.
Where does the PLC HMI license belong?
On the panel PC, because that is the system displaying the screens and running the client. Match the entitlement to the panel's platform and TwinCAT version before you deploy, and confirm it is accepted after installation. Installing the client files alone does not establish licensed operation, so verify the license state before you blame the deployment for a blank screen.
What changes if the panel PC runs Windows instead of Windows Embedded Compact 7?
The client, the license and the remote-client deployment apply the same way; packaging and startup differ. The CE-specific CAB package and shell command do not apply on Windows: install the client from the vendor's Windows installer and start it from an autostart entry instead. Confirm the TwinCAT version installed on the panel first, because package selection follows the panel platform and software build.