Run the floor. Connect the machines.

Andon Response Time: How Fast Is Fast Enough

Call to acknowledge, acknowledge to fix, two different clocks. What targets shops actually hit for each, and how to set one that survives contact with a busy floor.

First Pass Yield vs Scrap Rate: The Difference That Hides Rework

Two quality numbers that look similar and answer different questions. What rework hides from a scrap rate, and what each one leaves out on its own.

How to Get Operators to Actually Code Stops

Two taps, a list that fits on the screen, and one owner for the coding number. What gets a downtime reason picked instead of skipped.

How to Verify a Fix Actually Held

Before and after windows done right, the confounders that fake a win, and when it's fair to call an improvement zero instead of stretching to find one.

Minor Stops: The Losses Nobody Logs

The two minute stop that never makes it onto anyone's radar, why a plant needs thresholds instead of a human judgment call, and how the short stops add up past the breakdowns.

MTTR and MTBF for a Plant Floor, in Plant Words

What mean time to repair and mean time between failures actually measure, how much history makes them trustworthy, and what to do with a number built on too little of it.

Planned vs Unplanned Downtime: The Line That Changes Every Number

Where changeovers, PM windows, and breaks belong, and why moving that one line shifts your OEE, your Pareto, and your maintenance backlog all at once.

Running a Kaizen Event Off Machine Data Instead of a Guess

Pick the loss with a real dollar figure behind it, walk in with an evidence pack instead of a hunch, and measure the week after to know if it actually stuck.

Shift Targets People Actually Believe

How to set a real target from a routing instead of a guess, the mold time trap that wrecks a target without anyone noticing, and how to revise one without starting an argument.

The Shift Handoff That Actually Sticks

What the outgoing shift needs to write down, what the incoming shift needs to read first, and the five lines that matter more than a full report.

What a Golden Run Is, and Why It Beats a Nameplate Rate

The best sustained stretch a machine has actually run, as the speed reference. Why it beats a spec sheet number, and how drift away from it shows up.

Why Your Best Run Disappears

Setpoint drift, operator overrides, and tooling wear all erode a machine's best pace the same slow way. How to actually see it happening in the cycle trace.

6 min read

Is My OEE Good or Bad? What Other Shops Actually Run

World-class OEE is 85 percent. Most plants run closer to 60 percent. See the real OEE benchmarks, why your number probably looks low, and what to aim for.

9 min read

Downtime Reason Codes Operators Will Actually Use

Fewer codes beat more codes. What a usable downtime reason list looks like per vertical, a policy for stops too short to ask about, and coding at shift end.

8 min read

Your OEE Dropped Because You Started Measuring Everything

Automated OEE almost always reads lower than the clipboard number it replaces. Why that's the number to trust, and what your first accurate month is actually for.

8 min read

Tracking What Machines Run Without Work Order Ceremony

A lightweight way to tell a system what a machine is making without scheduling a full job. When paperwork earns its keep, and when it's just friction.

7 min read

Shift Comparison That Changes Behavior

D, E, and N shift analysis done right. Same machines, same mix, or the comparison lies. What to do when nights actually run worse.

9 min read

SPC and Cpk for Job Shops, Without the Six Sigma Seminar

Control charts and process capability explained plainly. What a Cpk below 1.0 actually costs, the spec limits nobody sets, and alerting on drift instead of reading charts.

8 min read

Questions to Ask an OEE Vendor Before You Sign

The measurement questions worth asking before a pilot: shift schedules, unmeasurable machines, read-only access, and what happens when the vendor is wrong.

7 min read

An Andon Call Is Only as Good as How Fast Someone Answers It

Call counts flatter a plant. Time to acknowledge tells the truth. What an andon system needs to actually work, and the metric that proves it does.

8 min read

Changeover Time You Can Actually Cut, Once You Stop Guessing at It

Changeover is planned downtime, not a failure. How to measure it accurately, rank it by dollars, and the first three moves that actually shorten it.

9 min read

Connecting Citizen and Star Swiss-Type Lathes: It's the Control, Not the Badge

Citizen and Star Swiss lathes carry different controls across their own lineups. How to find out which one is on your machine, and which guide actually applies.

9 min read

Monitoring a Machine With No Controller at All

Some machines never had a PLC and never will. How current-sensing, digital contacts, and cycle counting cover them without touching the machine's own wiring.

8 min read

How Many Machines Does One Gateway Actually Cover

The real answer to a question every quote depends on. Technical capacity, what limits it in practice, and the packaging most shops actually buy.

9 min read

Reading 4-20mA Sensors Through a Gateway

Current loops, loop power, and turning a raw ADC count into a PSI or a flow rate. What a clamp meter confirms, and what it can't tell you.

8 min read

Dry Contacts and Stack Lights as Machine State

A switch, a relay, a stack light segment, wired straight into the gateway's digital inputs. Debouncing a noisy contact, and what a bare on/off tells you.

8 min read

Wired, WiFi, or Cellular: Choosing a Gateway's Uplink

The uplink is the one network decision every install makes. Why wired wins by default, when WiFi is fine, and what cellular is actually for.

9 min read

Part Counts From a PLC Counter

Rollover, resets, and why a counter beats a derived cycle signal. What a counter tag's delta math protects against, and what to check on day one.

9 min read

What to Put on a Wallboard Operators Actually Look At

Most shop-floor TVs get ignored within a week. What earns a second glance, what's decoration, and how to avoid a wall of red nobody trusts.

8 min read

Tracking Scrap Without a Formal Quality System

You don't need SPC software or an ISO audit trail to know what's being scrapped and why. A lightweight way to start that a real quality system can build on later.

9 min read

What Each Machine Signal Means: Naming a Tag's Role

A counter, a run state, an alarm, a setpoint, a process value. Why the role you assign a tag matters more than what you happened to name it.

8 min read

Energy Monitoring: A First Win on Machines You Can't Connect Yet

Idle power waste is real money hiding in plain sight. How to find it with a power meter or a clamp-on sensor, before a full protocol integration is even on the table.

9 min read

Network Segmentation for Plant Monitoring

VLANs, the outbound-only gateway pattern, and what to put on the machine network. What IT will ask, and the direct answer to each question.

8 min read

What Happens When the Internet Goes Down

The gateway keeps polling and buffering locally, then backfills once the link returns. What the dashboard shows while a site is dark, and what's proven so far.

8 min read

What Is OEE and How Do You Calculate It?

The OEE formula explained plainly, with a worked example on one machine, the three factors that make it up, and why the textbook 85 percent target misleads more shops than it helps.

8 min read

TEEP vs OEE vs Utilization: Which Number Should You Use?

OEE, TEEP, and utilization answer three different questions. What each one measures, a worked example on the same machine, and which one to actually report to which audience.

9 min read

How Much Does Downtime Actually Cost? A Worked Cost Model

A downtime cost model a supervisor can redo on paper: lost throughput, labor paid to stand idle, expedite and overtime recovery, and why the sticker shock is usually accurate.

8 min read

Takt Time vs Cycle Time vs Lead Time, With Worked Examples

Three time measurements manufacturers mix up constantly. What each one actually answers, worked numbers on the same line, and how to spot which one is actually the problem.

9 min read

The Six Big Losses, Mapped to What Modern Monitoring Actually Sees

The classic TPM six big losses framework, checked against what a machine-connected system can actually detect automatically versus what still needs a human to say what happened.

9 min read

Preventive vs Predictive vs Reactive Maintenance: Real Costs and When Each Is Right

What each maintenance strategy actually costs, the data each one needs to work, and why most shops should run a deliberate mix instead of picking a single winner.

8 min read

How to Run a Daily Production Meeting Off Real Numbers

The 10 minute tier board meeting: what to review, in what order, why the same five minutes of status talk repeats every day without real numbers, and how to end with actions, not updates.

8 min read

Paper Travelers and Whiteboards vs Digital Tracking: What Actually Gets Lost

Paper and whiteboards are not the enemy of a good shop floor. What information disappears between them and a system of record, and where paper still legitimately wins.

8 min read

Machine Utilization for Job Shops: Why Spindle Time Isn't the Whole Story

Spindle-on time is easy to track and easy to misread. What utilization actually leaves out for a high-mix job shop, and the numbers worth watching alongside it.

9 min read

Bottleneck Identification: Theory of Constraints on a Real Production Line

How to find your line's actual constraint with data instead of a guess, why fixing a non-bottleneck wastes effort, and a worked example showing the constraint moving after a fix.

9 min read

What Is MES, and Do You Actually Need One?

MES, monitoring, and ERP explained plainly, with the clear boundary between them, a checklist for whether a full MES is worth the cost, and what to buy first if you're not sure.

8 min read

The Hidden Factory: The Capacity You Already Own

The term for the output a plant could produce with equipment it already has, if the invisible losses were fixed. What makes up the hidden factory, and how to find it before spending on new capacity.

9 min read

What a 30 Day Pilot Should Actually Prove

Written success criteria before day one, the two things a pilot has to show on a real floor, and what to walk away from when it can't.

9 min read

Machine Monitoring Pricing Models Explained

Per machine, per site, per user, hardware bundles, and activation fees, what each pricing model actually bills you for, with Spall's own pricing structure as the worked example.

8 min read

The First Dollar Finding: What Week One Usually Turns Up

What the first priced loss looks like once real data starts flowing, why it is usually a changeover or a minor stop instead of a big breakdown, and how to react to it.

9 min read

Build vs Buy: Pulling Your Own PLC Tags Into a Spreadsheet

What a homegrown PLC-to-spreadsheet or PLC-to-SQL setup actually takes to build, what breaks around machine ten, and what you keep either way.

8 min read

What to Show the Plant Manager in Week One

The three screens worth putting in front of a plant manager after the first week of real data, the one number that matters most, and what to hold back until later.

9 min read

Machine Monitoring ROI: A Worked Example

A twelve machine shop run through the loss and savings math on the ROI calculator's own assumptions: weekly losses, conservative savings, payback, and which two inputs actually move the answer.

8 min read

Why OEE Alone Does Not Change Behavior

A percentage nobody owns versus a dollar figure someone does. Why OEE on its own rarely drives a fix, and what turns the same data into action.

8 min read

What IT Will Ask About a Monitoring Gateway

The questionnaire a plant's IT department runs before a gateway touches the network: read-only, outbound-only, data egress, two factor, and what a good answer looks like.

8 min read

Choosing the First Line for a Pilot

Why the first machine to monitor should be repetitive with a PLC or counter already present and a leader who owns it, not the worst line on the floor.

8 min read

From Pilot to Plant: A Rollout Order That Holds Up

The sequence that works after a pilot proves itself: rates before dashboards, who owns coding a stop, and when the second gateway actually earns its place.

8 min read

What Verified Savings Actually Means

Before and after windows, attribution versus causation, and why a ledger that can show zero is worth more than one that can only show a win.

8 min read

How to Read a Vendor Case Study

What to ask for behind a percentage in a vendor case study: sample size, the measurement window, the basis for the number, and a checklist before you cite it.

~8 min

What Spall's AI Learns From Your Machines, and What It Does With It

A per-machine model forms automatically from telemetry, production, and downtime history already flowing in. What it learns, what it needs, and where it stays silent.

~8 min

Ask Your Factory: Twenty Questions a Plant Manager Asks, and What the Answer Cites

Plain-language questions about downtime, quality, shifts, and money, answered from live data with the sources named. What it can answer, and what it says when it can't.

~8 min

How Spall Predicts a Failure: From the Machine's Own History, With the Count Behind It

A failure-risk percentage built from a machine's own count of past breakdowns: MTBF, the odds of a failure this week, what it costs to be wrong, and when there isn't enough history to say.

~7 min

How AutoML Turns Your Best Run Into a Live Reference

The AI mines every completed cycle for the fastest sustained stretch a machine has run, then watches recent cycles against it continuously, with the chart and the dollar figure behind the finding.

~8 min

Alarm Sequences That Precede a Stop: How a Precursor Is Found and Turned Into a Trigger

How Spall mines a machine's alarm history for codes that reliably fire before a downtime reason or a scrap event, and turns a real pattern into a one-click trigger.

~8 min

AutoML on Any Controller: Learning Normal From Telemetry Shape, Not a Brand's Tag Names

How Spall decides which signals a model can fairly learn from, using what a value actually does over time instead of a tag name, a protocol, or a role a person forgot to set.

~7 min

AI Without a Data Scientist: What Needs Configuring, and What Needs Deciding

Spall's model selection, training, and scoring run with no setup screen at all. What a plant still has to decide: cost rates, reason codes, and which findings to act on.

~7 min

What the AI Needs From Your Plant: Data Readiness in a Week

Which models come online in a day, which need a full week, and which take longer, mapped against what a floor actually has to do to get there.

~7 min

How to Read an Insight Card: Evidence, Confidence, and the Blank That Means Something

Anatomy of an ML insight card: what the title claims, what the evidence backs it with, what the dollar figure means, and what a blank or a learning state is actually saying.

~8 min

From AI Finding to Verified Savings: The Loop That Proves It

Linking an insight to the action taken opens a measurement window. What comes out the other side: verified, no change, or worse, priced only against a real rate, confounders named either way.

8 min read

Energy Cost per Part: Reading Power Draw Against Production

How to turn a machine's metered kWh into a real dollar figure per part, why idle time skews the number, and what the Energy page prices today.

9 min read

A Wallboard a Crew Reads from Twenty Feet

The ten tile types, TV-scale type, and the fit-to-tile rule that keeps a wallboard from clipping a row or freezing a stale number.

8 min read

Shift Attribution: Getting a Stop on the Right Shift

How Spall decides which shift a stop or a part belongs to, why cross-midnight and rotating patterns trip it up, and what a wrong schedule does to every downstream number.

9 min read

The Preventive Maintenance Calendar the Data Keeps

PM tasks on a real cadence, what overdue actually means, MTBF as the interval a task should run on, and how the calendar connects to the repair queue.

~7 min

Connect Your AI Assistant to Spall With MCP

Point Claude Desktop, Cursor, or Copilot at your Spall tenant with one package and a read-only key. What each tool reads, what it never does, and the ask scope and budget.

~6 min

Your Plant in Claude, Copilot, or Cursor: What a Read-Only MCP Server Does, and What It Never Does

Why a plant would let its own AI assistant read Spall, what fourteen read-only tools return, the basis every answer carries, and the three things the package refuses to do.