Andrei Fogoroș

Huge appetite for #CyberSecurity, #SoftwareCompliance, #FOSS, #SmartManufacturing, #BigData, #IIoT, #DigitalTransformation, #Industry40, #soccer, #traveling.

[IIoT World @IIoT_World]

More hardware on the shop floor means more attack vectors. Audi's Edge Cloud 4 Production centralizes all apps on one software foundation, shrinking the cybersecurity surface of the plant. Partner content with @Siemens. #sie_us

[IIoT World @IIoT_World]

Model Context Protocol (MCP) replaces fragile prompt-based tool instructions with strict input/output schemas, so AI in industrial systems can discover and interact with available capabilities in a structured way.

[Lucian Fogoros @fogoros]

IIoT World's Industrial AI Summit opens September 9-10 as a free virtual conference covering industrial AI architecture, predictive analytics, and operational intelligence.

[Lucian Fogoros @fogoros]

Data governance programs that celebrate nothing sustain nothing. When a targeted data improvement produces visible results, a better report, more reliable decisions, or higher data accuracy, and someone asks what else governance can fix, momentum is building.

[IIoT World @IIoT_World]

Just published on @iiot_world: factory documents block AI agents because standard operating procedures, quality records, and engineering documents contain tables, revision histories, signatures, embedded images, handwritten notes, checkmarks, diagrams, scanned pages, tolerances, symbols, and embedded CAD references that AI cannot use without decomposition first. Partner content with Adlib. #sie_us #adlib_iiot

[Lucian Fogoros @fogoros]

Adoption spreads faster through a small number of people on the production floor who have seen predictions confirmed on their own equipment than through any company-wide training program. Partner content with TDK SensEI. #TDK_iioT

[Lucian Fogoros @fogoros]

Crane monitoring is moving from periodic sampling toward always-on, asset-specific architectures. Partner content with @AI4ProdOutcomes. #Infinite_iiot

[Carolina @CRudinschi]

Standard PdM was built for steady-state machines at constant speed. A duty-cycle crane accelerates, decelerates, and shock-loads hundreds of times a day under extreme heat. Two completely different monitoring problems. sponsored by @AI4ProdOutcomes #infinite_ai

[Lucian Fogoros @fogoros]

Manufacturers cannot solve brownfield AI problems by replacing every legacy system. Partner content with Adlib. #sie_us #adlib_iiot

[Lucian Fogoros @fogoros]

You are the reliability manager. When the main hoist on the melt-shop crane drifts toward a seizure at 2 AM, you get the call. A failed hoist does not idle one machine, it idles the entire bay. Partner content with @AI4ProdOutcomes. #Infinite_ai

[Lucian Fogoros @fogoros]

Ucaneo uses electrochemical Direct Air Capture. Siemens is deploying the Siemens Xcelerator portfolio as the automation and digitalization backbone, spanning from the first-of-a-kind Berlin plant all the way through to commercial scale-up. Partner content with @Siemens. #sie_us

[Carolina @CRudinschi]

What happens when your AI model doesn't know that a vibration spike during the first 10 minutes after a product changeover is mechanical settling, not bearing wear? Partner content with TDK SensEI. #TDK_iioT

[Carolina @CRudinschi]

At a leading steel producer, a melt-shop crane hoist bearing was drifting toward seizure. Continuous monitoring caught it 90 days out, they scheduled a re-lube, and 4 hours of unplanned downtime never happened. sponsored by @AI4ProdOutcomes #infinite_ai

[Carolina @CRudinschi]

Predictive maintenance is not just an equipment problem anymore. Read more (link in the comments). Partner content with TDK. #TDK_iioT

[IIoT World @IIoT_World]

How many of your crane's 300-400 lift cycles per shift does your PdM sensor actually see? Battery sensors wake up a few times a day and go quiet during load. The answer is close to zero. sponsored by @AI4ProdOutcomes #infinite_ai

[Lucian Fogoros @fogoros]

Return on Mitigation is calculated using the Cyber Risk Equation: Risk equals Threats multiplied by Vulnerabilities multiplied by Consequences. You convert that to Annualized Loss Exposure by multiplying event probability by estimated consequence cost.

[Lucian Fogoros @fogoros]

AI-driven 3D reconstruction of utility poles runs as a three-stage pipeline. Field crews capture multi-angle imagery in roughly two minutes, significantly reducing time spent in hazardous roadside environments.

[Lucian Fogoros @fogoros]

Automation programmer shortage slows US reshoring and new plant construction. Siemens Eigen Engineering Agent uses AI to design, deploy, and troubleshoot PLC and DCS projects. Partner content with @Siemens. #sie_us

[Lucian Fogoros @fogoros]

Esther Stepansky's S4x26 framework signals a shift in how OT security governance needs to work. Some maintenance debt can be fixed immediately. Some must be managed over time. Some must be accepted because the cost of remediation exceeds the risk.

[Carolina @CRudinschi]

Smaller manufacturers without a dedicated data science team are better served by purchasing a purpose-built predictive maintenance solution than building one. Partner content with TDK. #TDK_iioT