Body, Paint and Final on One Edge Stack
What to connect in body, paint, and final assembly, how to move it without opening the plant network, and how to run torque and weld models on the same node as SCADA.
19 minute read PDF
One paper for each use case on this site. Each starts with a self-check for your plant, shows what to connect and how to keep the plant network closed while you do it, names where these projects go wrong, and ends with what to do Monday. Every number is sourced.
What to connect, how to move the data without opening the plant network, where these projects go wrong, and a phased plan to start Monday.
What to connect in body, paint, and final assembly, how to move it without opening the plant network, and how to run torque and weld models on the same node as SCADA.
19 minute read PDF
A path from PC sprawl to one console: inventory the line, read PackML state into a common model, move HMIs one at a time, and replace per-skid modems with one logged path.
22 minute read PDF
For SMT process and test engineers: joining machine data by board and time, cutting AOI false calls, catching paste defects at the printer, and protecting customer program IP.
22 minute read PDF
How to read CIP skids, fermenters, and fillers without changing their programs, and build the CIP, batch, and lot genealogy records QA signs and FDA asks for.
21 minute read PDF
For chemical plants and blenders: a read-only layer that watches gas detectors and the SIS, keeps OSHA-ready test records, scores alarm load, and works at sites with no network.
23 minute read PDF
How sterile manufacturers connect particle counters, pressure, and BMS data on one clock, alarm at the edge, and map Part 11 and Annex 11 controls onto the design.
23 minute read PDF
SCADA for remote solar and wind sites: store-and-forward at the edge, separating comms gaps from real outages, brokered OEM access, and CIP-003-9 in practice.
21 minute read PDF
For utilities: what substation condition data is worth collecting, an edge architecture that never touches protection, a station-by-station rollout, and how NERC CIP applies.
21 minute read PDF
For production and SCADA teams: closing exposed pad devices, keeping API 21.1 records through outages, routing pumpers by exception, OOOOb evidence, and patching without a truck roll.
23 minute read PDF
A practical plan for utilities: close internet exposure first, put an edge node at each site, route lift station alarms with a record, and build compliance reports from the historian.
24 minute read PDF
A staged program for distribution and NRW teams: fix the AWWA audit, tighten one DMA, watch night flow, measure and manage pressure, and capture the transients ordinary loggers miss.
22 minute read PDF
How cement plants fix analyzer, feeder, and lab timing data first, then run soft sensors and an advisory optimizer on an edge node beside the DCS, with the operator in charge.
21 minute read PDF
Build one millisecond record of every stand with the equipment the mill already owns, use it for cobble and gauge reviews, and fence level 1 off from the office with zones and conduits.
22 minute read PDF
For growers and irrigation crews: what to measure at the pump first, which radio fits which job, a private network on towers and field nodes, and the Plow Horse farm twin.
23 minute read PDF
What to measure on a club car, how to move data from car to garage over sub-GHz Wi-Fi with store and forward, and how to keep it private in a shared paddock.
23 minute read PDF
Vision, robotics, scheduling, and maintenance: what each one needs from the edge, and how to put it into production safely.
Vision inspection on the line: choosing rules or deep learning, lighting and labels, the PLC reject handshake, measuring the station like a gauge, and handling drift.
21 minute read PDF
Where VLA models fit beside the robot controller, what the latency budget allows, how to collect your own demonstrations, and how to test a cell before trusting an autonomy claim.
24 minute read PDF
Why ERP plans go stale by mid-morning, what a finite schedule needs from the floor, and how to reschedule on machine, material, and order events without making the floor nervous.
21 minute read PDF
A floor-level AI assistant for maintenance techs: cleaning work orders and manuals, cite-or-refuse retrieval, LOTO limits, and a phased rollout that runs on site.
22 minute read PDF
A method for pricing the same outcome on every quote, with every input labeled by where it comes from.
A line-by-line price comparison of EmberNet, ZEDEDA, and Litmus at the same redundant, segmented outcome, with every input labeled by source and a five-step method for checking any quote.
23 minute read PDF
Tell us what runs on your floor. An engineer will tell you how EmberNET would run it, and what we would connect first.