Edge intelligence for industrial equipment
KNOCE combines rugged control hardware, on-device AI, and a cloud monitoring platform to catch equipment failures before they stop production — proven on frac pads, built for any plant.
Proven where it's hardest
We didn't build KNOCE in a lab. We built it for hydraulic fracturing — remote sites, brutal duty cycles, no fiber, no second chances. If it survives a frac pad, your plant is easy.
The platform
Data from any equipment — direct I/O or beside your existing Siemens, Allen-Bradley, and CODESYS systems.
Live dashboards, fleet health scores, and severity-tiered alarms across every site.
AI anomaly detection with failure modes, confidence scores, and remaining-useful-life estimates.
Detections become work orders automatically — parts, labor, and telemetry attached.
KNO1-PLC hardware
Run it as the full controller — or as a smart gateway beside the PLCs you already own, reading data over OPC-UA and touching nothing.
For integrators & service providers
License the platform, lease the hardware, and launch a recurring-revenue monitoring business under your own brand.
Frac Operations Platform
A frac supervisor or remote operations center has never had a single credible screen showing the entire spread — every pump unit, every genset, and the stage in progress. The OEM's HMI shows one unit at a time; everything else lives in spreadsheets and judgment. KNOCE Frac Operations is that screen.
The four numbers on top of every shift
Representative values. Idle fuel burn deserves the last word: turbine fuel consumed between stages has always been invisible on every other system. KNOCE prices it live, per hour, so the cost of an idle decision is on the screen while the decision is being made.
See the spread
A tile per pump unit — status, health, and live vitals — with the genset power strip, the current stage panel, and an alarm ribbon across the top. The screen a supervisor keeps open all shift.
Any unit, one click deep: discharge and suction pressure, rate, RPM, power-end and fluid-end vibration, bearing and lube-oil temperature — live and trended — beside run-time counters and service-interval status.
Per-genset load curve with droop characterization, heat rate against the ideal curve, fuel rate, and exhaust temperature — the difference between a healthy turbine and a fouling one, visible on one chart.
The live stage panel holds targets against actuals — rate, pressure, proppant, fluid volume, elapsed time — so the spread view and the stage plan are finally the same screen.
Run it with discipline
Not a scrolling wall of red. Alarms carry priority (critical, high, medium, low, journal) and a full lifecycle — active, acknowledged, cleared — with shelving and state-based suppression, so nuisance alarms stop training your operators to ignore the board.
Operational requests — reduce rate, idle a unit, bring one online, hold — move through a five-step flow, every step timestamped and attributed:
Remote operations with a paper trail — the record of who asked, who approved, and what happened exists before anyone asks for it.
Maintain on evidence
Fluid-end washout. Power-end bearing wear. Compressor fouling. Each finding arrives with a severity, a confidence percentage, and the trend evidence that triggered it — so an engineer can verify the reasoning, not trust a black box. One click turns a confirmed finding into the right work order: a washout becomes a fluid-end change, bearing wear becomes a bearing inspection.
Fluid-end hours, power-end hours, and total barrels pumped accumulate live on every unit — and when a counter crosses its service limit, the work order raises itself. Orders flow open → assigned → in progress → complete, with a technician completion view built for gloved hands, and every completed job feeding the unit's permanent history.
KNOCE Frac Operations is a monitoring and supervisory-request layer. All protective functions — emergency shutdown, overpressure protection, interlocks — remain with the certified hardware layer, fully independent of this software. That boundary is deliberate, it's how responsible operators expect the stack to be built, and we put it in writing.
Under the hood
Data arrives over MQTT Sparkplug B, matching pad broker topology — from KNO1-PLC gateways or existing acquisition.
discharge_psi · suction_psi · rate_bpm · rpm · power_end_vib · fluid_end_vib · bearing_temp · lube_oil_temp · hhp_used · fluid_end_hours · power_end_hours · total_pumped_bbl · load_kw · load_pct · heat_rate_btu_kwh · fuel_rate_gph · exhaust_temp · stage_targets · proppant
Platform
Most monitoring stacks are a science project: sensors from one vendor, historian from another, analytics from a third, and a CMMS that never hears about any of it. KNOCE is one platform from the I/O terminal to the work order.
01 · Connect
The KNO1-PLC acquires signals directly through 76 on-board I/O channels, or reads from the control systems you already run — no rip-and-replace, no proprietary sensor lock-in.
Plain MQTT moves bytes; Sparkplug B moves meaning. Every tag arrives in the platform already named, typed, and stateful — so a new site comes online in hours, not integration-weeks. It's the difference between plumbing and a nervous system.
02 · See
Every asset across every site with a live health score, status, trend sparkline, and gateway connectivity — filterable by site, customer, or condition.
Built to ISA-18.2 practice: priorities, acknowledge with attribution, shelving, and state-based suppression — plus source attribution, so you always know whether a threshold, the cloud AI, or the on-device AI raised it.
A purpose-built monitoring-center screen: priority attention queue, unacknowledged alarms, and the live AI detection feed — for your team or a service provider's NOC.
03 · Predict
AI predictions are only as good as the data behind them. KNOCE models sharpen with every operating hour on your equipment — and every detection shows its confidence and its evidence, so your engineers can verify the reasoning instead of trusting a black box.
04 · Act
A confirmed detection creates a work order with the failure mode, recommended action, suggested parts, and labor estimate — pre-filled, not blank. Run-time counters raise routine orders the same way: when fluid-end hours cross the service limit, the order writes itself.
Work orders live in Jira with equipment-level organization and two-way sync. Teams already on Jira change nothing; teams without a CMMS get one that works on day one.
Every alarm, detection, command, and completed job accumulates per asset — with a technician completion view built for the field, so the record gets written where the work happens.
Operational requests move through request → validate → acknowledge → execute → audit — every step timestamped and attributed. Remote operations with a paper trail, and a hard boundary: protective and safety functions always remain with the certified hardware layer.
Edge AI
Most platforms ship raw data to the cloud and think there. KNOCE thinks at the machine — on acceleration built into the controller — and uses the cloud for what the cloud is good at: fleet learning, history, and heavy analysis.
TPM 2.0 hardware root of trust and an IEC 62443-ready design. In gateway mode, integration is read-only — your control logic is never touched, and the attack surface a new device usually adds simply isn't there.
KNO1-PLC
One sealed industrial controller that replaces the PLC, the I/O rack, and the cabinet — or drops in beside your existing systems as a smart gateway. Adaptive compute, real-time control, and AI acceleration in a single 12-inch box.
New builds and upgrades: the KNO1-PLC is the control system — deterministic loops, integrated I/O, and AI in one enclosure.
Existing plants: the KNO1-PLC sits beside your control systems, reads over OPC-UA or Modbus, and adds AI without touching a rung of logic.
Which mode? If you're building or replacing a panel, Mode 1 gives you control and intelligence in one purchase. If the plant already runs and the PLCs are fine, Mode 2 adds the intelligence layer with zero disruption — and you can migrate to Mode 1 asset by asset whenever a panel comes due.
System specifications
| Compute module | AMD Xilinx Kria K26 SOM — quad-core Arm Cortex-A53 @ 1.33 GHz application processor, dual-core Cortex-R5F @ 533 MHz real-time processor, AI acceleration up to 1.4 TOPS INT8 |
| Memory & storage | 4 GB DDR4 (64-bit @ 2400 Mb/s) · 16 GB eMMC · 512 Mb QSPI flash |
| Operating system | Linux Ubuntu Server with full development stack |
| I/O | 76 configurable channels: 20 high-power (5–36 VDC, 12 A per channel) + 56 general-purpose (4-20 mA, 0-25 mA, 0-5/0-10 VDC, resistance 400 Ω–4 MΩ, frequency, sub-1 VDC with microvolt resolution) |
| Analog front end | Precision FET-input instrumentation stage, 0.5 mV max offset, industrial grade |
| Communication | 2× LAN 10/100/1000 (3rd optional) · RS232 · RS485/RS422 · CAN — all on sealed M12 connectors |
| Programming | IEC 61131-3 (LD, FBD, ST, SFC) via native CODESYS · Python · C · C++ |
| Agent interface | Model Context Protocol (MCP) server on-device — exposes I/O, diagnostics, alarms, and setpoints to Claude and other MCP-compatible agents |
| Security | TPM 2.0 hardware root of trust · IEC 62443-ready |
| Environmental | IP69 ingress rating · −40 to 185 °F (−40 to 85 °C) system rated · sealed M12 throughout |
| Power | 9–36 VDC input (24 VDC @ 5 A) · ~15 W typical draw |
| Envelope | 12" × 10.15" × 1.74" — panel or surface mount |
Built different
Legacy PLCs execute logic; the KNO1-PLC also runs inference. Anomaly detection happens on the same device reading the sensors — no external edge server, no round trip.
Linux with Python and a full toolchain means modern software runs at the machine — connectors, agents, and analytics that proprietary firmware can never host.
CODESYS instead of a locked vendor IDE. Standard protocols instead of a captive ecosystem. Your integrator can work on it; your team isn't hostage to one vendor's licensing.
Industries
The same platform that watches a turbine at 30,000 operating hours watches a chiller, a mill, or a substation. If it has sensors and it can't go down, KNOCE fits.
Turbine-driven frac fleets run some of the hardest duty cycles in energy — at sites with no fiber, extreme temperature swings, and zero tolerance for surprise downtime. KNOCE was engineered here first: deterministic turbine and pump control on real-time cores, with AI watching the same signals for combustor distress, bearing wear, vibration signatures, and fuel-system anomalies.
On-device intelligence means detection continues when the satellite link doesn't. Alerts queue locally, sync on reconnect, and escalate to the fleet view the moment they land.
Continuous health monitoring for the assets grids depend on — with substation-grade environmental ratings and offline detection for remote sites.
Fewer surprise stoppages on the equipment that carries production, with maintenance planned on evidence instead of calendars.
Pump stations and treatment plants — including the remote ones where connectivity is a rumor and a failed pump is a public problem.
Vibration-driven wear detection on machines that grind through everything — catching bearing and gearbox degradation weeks before the seizure.
Comfort and uptime are the product. Monitor the mechanical plant behind the terminal, and the backup power behind everything.
The platform is equipment-agnostic by design: if it produces a signal, KNOCE can learn its normal and flag its abnormal. Tell us what you run.
Partner program
System integrators, controls contractors, and industrial service companies use KNOCE to launch their own 24/7 monitoring and predictive maintenance business — under their brand, on our stack, without building a software company first.
The opportunity
Integration work is project revenue: it ends when the commissioning does. A monitoring service is a relationship: every site you connect pays you every month, and your value compounds as the AI learns your customers' equipment. You already have the two hardest assets — the customer trust and the site knowledge. KNOCE supplies the rest.
The full stack — dashboards, AI detection, alarms, work orders, and the 24/7 operations view — delivered as your service, with your name on the customer relationship.
Lease KNO1-PLC gateways instead of buying hardware up front. Low entry cost per site keeps your service margin healthy from the first customer.
Deployment training for your engineers, L2/L3 platform support behind your L1, and co-branded material to sell with. You run the service; we keep the platform solid.
How it works
We map your customer base and territory together, and agree what exclusivity you can earn.
Deploy a gateway at a friendly customer. Data flows in days; the first AI baselines follow.
Your brand, your pricing, your customer contract — running on the KNOCE platform with our support behind you.
Hit your growth commitments and your territory is yours. Real exclusivity for partners who perform.
Questions partners ask
No. In gateway mode the KNO1-PLC reads over OPC-UA or Modbus — read-only. Your customers' control logic, safety systems, and vendor warranties stay exactly as they are. That's deliberate: it makes your sale easier and your risk smaller.
You do. The service contract, the pricing, and the invoice are yours. KNOCE is your technology supplier, and registered accounts are protected in the partner agreement.
If your engineers commission SCADA and PLC systems today, they already have the hard skills. Gateway deployment is protocol configuration, not software development — and we train your first team hands-on during the pilot.
Your team handles first-line customer questions; KNOCE engineering handles platform issues behind you with defined response times. Your 24/7 promise is backed by our stack and our support commitment, in writing.
The standard program is co-branded — your service, "powered by KNOCE." Full white-label is available at higher commitment tiers. Either way, your customers see your company as their monitoring provider.
Your customers' raw operational data belongs to them, and identifiable data can respect residency requirements. De-identified telemetry improves the AI models that make your service better every month — that's the flywheel your customers benefit from.
If you install, integrate, or maintain industrial control systems, you're already most of the way to a recurring-revenue monitoring business.
Company
KNOCE was founded in Houston, Texas at the intersection of two disciplines that rarely share a room: industrial operations software and rugged control hardware. We build for the people who get the 2 a.m. call when a machine goes down — because we've taken that call.
KNOCE Systems is a Houston, Texas industrial technology company that builds AI-powered equipment monitoring and control. Its products are the KNO1-PLC, a sealed edge controller with on-device AI that works as a full IEC 61131-3 controller or a read-only smart gateway beside existing Siemens, Allen-Bradley, and CODESYS systems, and the KNOCE monitoring platform, which turns live telemetry into ISA-18.2 alarms, evidence-cited AI diagnostic findings, and automatic work orders. KNOCE's flagship application is turbine-powered hydraulic fracturing operations, and the same platform serves power generation, manufacturing, water, mining, and facilities. KNOCE is part of the Knomatic family of companies.
What we believe
Cloud-only monitoring fails exactly when you need it most — when the connection drops at a remote site. Detection has to live on the device, with the cloud adding fleet-scale learning on top.
Dashboards don't fix bearings. Every detection in KNOCE carries a failure mode, a recommendation, and a work order — because the point is the repair, not the chart.
IP69, −40 to 185°F, sealed M12 connectors. Industrial equipment doesn't live in climate-controlled racks, and the hardware watching it shouldn't need to either.
CODESYS over captive IDEs. Standard protocols over proprietary ecosystems. Your integrator can service it, your team can extend it, and your data is yours.
Heritage
KNOCE is a Knomatic company — built on the operations-software platform Knomatic develops for industrial, logistics, and fleet businesses, and paired with control hardware engineered for the energy sector's hardest environments. The name carries the mission: KNO for knowledge, CE for the control engineering it's built on. Know your machines.
Contact
Three ways in, depending on who you are. Every message reaches an engineer, not a queue — expect a reply within one business day.
For operators and plant teams: a walkthrough of the platform on equipment like yours — fleet view, AI detections, and the work-order loop, live.
Request a demoFor integrators and service companies: a conversation about your market, your customer base, and what a monitoring service could look like under your brand.
Start the conversationFor the technical questions: protocols, deployment modes, security posture, spec sheets, and everything the marketing page didn't answer.
Ask engineeringKNOCE Systems · Houston, Texas · knocesystems.com