Where OT, IT and finance meet. Business-focused insights on secure remote access to industrial networks: reducing downtime, truck rolls and vendor risk.
Hot take from the plant floor: "one remote access platform for everything" is how you break operations.
Headless automation needs strict policy + least privilege.
Screen workflows need capacity + rate-aware monitoring.
Same thresholds != same risk. #ICSsecurity#RemoteAccess
At @@IIOTaccessLog we obsess over the unsexy part: operating remote access under uptime pressure.
Headless automation != screen sessions.
Different rhythms, different risk, different logging.
Build dual control planes, then give OT/SOC one view. #OTsecurity#IIoT
Problem: remote viewing helps, but the operator still has to walk into the hazard zone to act.
Solution: design remote access for true operation:
- embed in the native UI
- limit command paths
- define takeover rules
- test under low bandwidth
Contrarian take: "air-gapped = safe" often just moves risk into people and process.
When conditions are hazardous, you need controlled remote operation:
- authorized stations
- time-bound access
- audit trail
- fail-safe priority
Security isn't only disconnection.
Operational ROI framing I use for remote access:
MTTR ↓ + fewer truck rolls + fewer "hazard hours" onsite.
If you can't quantify:
- downtime cost/hr
- vendor wait time
- travel/escort time
- audit time
...you'll buy tools, not capability.
Remote access in high-risk ops isn't "screen sharing." It's a control surface.
- Embed it where work happens (HMI/nav UI)
- Constrain who can send commands
- Add priority + fail-safes
If it can't run safely in bad conditions, it's not ready.
My "remote to an HMI" checklist:
- Is the asset inventoried + owner named?
- Which role is allowed (view vs control)?
- Time-boxed approval path?
- Session recorded/logged?
- Break-glass procedure tested?
If you can't answer fast, pause.
Every time a new "remote ops" initiative hits the plant, the same gap shows up:
Tools get bought.
Procedures don't get built.
Remote access fails when it's a project, not an operating model.
Start with: roles, windows, logging, and ownership.
If you're evaluating an OT remote access platform, test it like operations:
- stable sessions (no weird disconnects)
- RBAC with view/control separation
- time-bound access + approvals
- full audit trail + easy exports
If it can't pass an outage window, it won't scale.
Problem: legacy embedded VNC on an HMI "works," until an audit or incident.
Solution:
- central gateway/jump host
- MFA + role-based access
- time windows tied to work orders
- session logging/recording
Keep the usability. Add control.
A quick ROI lens for OT remote access:
Minutes saved per vendor/engineer visit x frequency x loaded labor + avoided downtime.
If you can't quantify:
- truck rolls avoided
- MTTR reduction
- audit hours saved
...it's just screen sharing with risk.
Remote access to HMIs/PLCs isn't an IT "convenience"-it's an ops workflow.
If it doesn't:
- protect uptime/safety
- stay deterministic
- follow a runbook
...you'll create more downtime than you save.
Treat it like SCADA: engineered + governed.
Thread: retiring OSS VNC without breaking production.
1) Treat remote access as a reliability system.
2) Do it in rollout waves (site/region).
3) Build ROI around downtime avoided + truck rolls avoided.
4) Consolidate to 1 supported platform with audit.
What would you add? #OT
If you're rolling out remote access to PLCs/HMIs across multiple plants:
What's harder right now?
1) vendor governance
2) technical standardization
3) audit trail
4) operations buy-in
I'll share the playbook I use.
Legacy VNC cleanup checklist I use:
- inventory where VNC actually runs (HMI, eng laptop, test rig)
- map vendor dependencies
- find shared creds
- document ports/routes
- define "allowed" use cases
If you can't see it, you can't retire it. #ICS#OT
Remote access in OT isn't a convenience feature.
It's a reliability system and a security boundary.
If you can't answer "who accessed what, when, and why," you don't have remote support-you have unmanaged exposure.
If you're still on Ultra/Tight/TigerVNC in production, what blocks migration?
A) downtime windows
B) vendor lock-in
C) no owner/process
D) "it works, don't touch it"
Reply with your constraint - the plan changes based on which one is real. #OT#IIoT