.png){kind=logo}
Yes. Browser-based dashboards work on Chrome, Safari, Edge, and Firefox on any OS.
Real-time systems typically update dashboards every 1–5 seconds.
Yes. Enterprise views allow executives to see data from multiple factories in a single login.
Yes. You own your data. You can export your historical records to CSV/SQL before cancelling.
Caddis is designed for "High Mix / Low Volume" environments where setup reduction is critical.
No. It feeds accurate data into your MES (Manufacturing Execution System) so the MES isn't relying on manual input.
SCADA is for controlling industrial processes (turning valves); Machine Monitoring is for analyzing productivity.
ERPs schedule work, but they don't track if the work is actually happening in real-time. You need both.
Free tools are usually manual calculators. They cannot track real-time downtime or alert you when a machine stops.
Built-in OEE only sees that machine. It cannot aggregate data across a mixed fleet of 10 different brands into one dashboard.
Amper uses exclusively electrical sensing; Caddis offers electrical plus direct PLC/Digital I/O integration for deeper data.
Very little. Machine data packets are tiny (kilobytes), so it won't slow down your office internet.
Caddis is often preferred for SMBs needing faster, simpler deployment, while MachineMetrics targets enterprise-level high-frequency data.
(For Caddis) Yes, an open API is available for custom integrations with BI tools like PowerBI or Tableau.
Yes. Most providers have mobile-responsive sites or dedicated apps for iOS and Android.
You will need to reconfigure the gateway/IoT device, usually via a USB connection or AP mode.
Most hardware kits have LED indicators that blink when the sensor is triggered.
Yes. "Self-install" kits are designed for maintenance teams to install without external contractors.
Ensure the tablet is on the same WiFi network as the gateway, or check if the machine power is off.
Check the "debounce" setting. If the sensor is too sensitive, one part might trigger two counts (double counting).
Red usually indicates the machine is in a "Down" state or has exceeded a downtime threshold.
Yes, if you create a specific "Rework" button for operators to log when they are fixing a bad part.
The difference in time between the fastest and slowest cycle for the same part, indicating process instability.
Yes. You can monitor the "Active" signal from a Universal Robot or Fanuc arm.
Yes. CSV/Excel export is a standard feature for custom analysis.
Discrete monitors distinct items (parts/widgets); Process monitors continuous flow (gallons/liters), though Caddis focuses on Discrete.
If equipped with power monitoring sensors, it can detect spikes that indicate motor failure or a crash.
Grouping similar parts together in the software to analyze average cycle times across a product line.
Yes. Dashboards allow you to compare Day Shift vs. Night Shift utilization to balance workloads.
Yes. Swiss lathes (like Citizen or Star) are excellent candidates because they run high-volume cycles where seconds matter.
You can create a "Warm Up" downtime category so this time doesn't negatively impact the operator's performance score.
Running machines unattended (usually overnight). Monitoring is critical here to alert you remotely if a machine crashes.
Yes. Most systems allow SMS or Email alerts for specific triggers (e.g., "Machine 4 down for >15 mins").
Yes. You can set alerts (e.g., "After 500 parts, warn operator") to change tools before they break.
Yes. Additive manufacturing equipment can be monitored via API or by tracking extruder temperature/current.
It tracks when the machine stops. If management uses this to police bathroom breaks, the implementation will likely fail due to culture issues.
Offer bonuses for hitting utilization targets or for high data-accuracy (categorizing all downtime).
Yes. Giving operators visibility into their efficiency usually boosts performance through self-correction.
Many vendors (like Caddis) offer pilots or risk-free trials to prove value on a small set of machines.
A visual display (TV) on the shop floor showing real-time status (Green=Running, Red=Down) to alert supervisors to issues.
Modern interfaces are simple (like an iPad app). Training typically takes less than 15 minutes per operator.
Yes. TV dashboards (Andon boards) are recommended to gamify production and keep teams aligned on goals.
The system will record the stop as "Uncategorized." Supervisors can edit it later, or the system can prompt the operator when they return.
No. Automated counting removes the need for manual paper logs, actually reducing administrative work for operators.
Involve them early. Show how the data proves they need new tools or maintenance support, rather than using it for punishment.
No. It monitors the machine's health. The goal is to identify barriers (no material, bad tools) that prevent operators from succeeding.
Yes. Accurate cycle time data allows you to quote based on reality, not estimates, protecting margins.
Yes, machine monitoring is scalable. You can start with one bottleneck asset and expand later.
It varies, but in automotive/aerospace, downtime can cost $500 to $10,000+ per hour depending on the bottleneck.
Usually no. DIY solutions (Arduino/Pi) lack security, support, and scalable software, costing more in engineering hours long-term.
Yes. Most modern systems have APIs to push job data to ERPs like JobBOSS, Epicor, or E2 Shop System.
It allows "lights out" or unattended machining, letting one operator run more machines, effectively reducing labor cost per part.
It reduces overtime costs, delays new machine purchases (CapEx) by finding hidden capacity, and lowers scrap rates.
Costs typically range from $40–$100 per machine/month for SaaS, plus a one-time hardware fee ($200–$800).
Most shops see ROI in <3 months by recovering 10-15% of lost capacity through better visibility.
Yes, if you use hardware capable of monitoring amperage/voltage (like current transducers) alongside cycle data.
The theoretical fastest time it takes to produce one part (often the machine's nameplate speed).
Use monitoring data to benchmark current setup times and identify which changeovers take longer than the average.
It tracks actual runtime hours or cycle counts to trigger maintenance alerts based on usage rather than calendar dates.
Total Effective Equipment Performance (TEEP) measures OEE against 24/7/365 availability, not just scheduled shifts.
This usually means your "Ideal Cycle Time" is set too high (slower than the machine is actually running).
Usually no. Scrap reasons (e.g., "scratch," "dimension off") require human input, though the count can sometimes be automated.
World Class OEE is 85%, but the average manufacturing OEE is often around 60%.
Standards vary, but typically Setup is considered "Planned Downtime" and excluded from Availability, or tracked separately to improve SMED.
Planned downtime is scheduled (lunch, maintenance); Unplanned downtime is unexpected (breakdowns, missing tools).
Availability = (Run Time / Planned Production Time) × 100.
A short stoppage (usually <2 minutes) that is too short for an operator to log but accumulates to significant lost time.
Operators typically use a tablet/interface to select reasons (e.g., "Tool Broken," "No Material," "Setup") when the machine stops.
Quality = (Good Parts / Total Parts Produced) × 100.
Performance = (Ideal Cycle Time × Total Count) / Run Time.
Yes. You can tap into the pressure switch or cycle start relay to count strokes and track run time.
The system will log the disconnect. High-end hardware may have battery backups to send a "power loss" alert before shutting down.
No. SaaS (Software as a Service) solutions host the data in the cloud, eliminating the need for expensive local servers.
Non-invasive hardware kits (like Caddis) can typically be installed in 15-30 minutes per machine.
OPC Unified Architecture is a machine-to-machine communication protocol for industrial automation, common in newer PLCs.
Yes. Mazak machines often use the MTConnect adapter or the Mazak SmartBox for easy integration.
Yes. Haas machines effectively support monitoring via their Next Generation Control (NGC) or older serial ports.
Typically only outbound port 443 (HTTPS) is required, meaning you don't need to open dangerous inbound ports.
Install a current transducer (CT clamp) on the spindle motor or a vibration sensor to detect when the machine is cutting material.
Yes. Reputable providers use bank-level encryption (AES-256) and do not control the machine, only read data from it.
An input/output device that translates electrical signals (like a light turning on) into digital data the software can read.
Yes, but hardwired Ethernet is recommended for factory floors due to electrical interference that can disrupt WiFi signals.
Most modern systems are cloud-based and require internet, but many offer "store and forward" capabilities if the connection drops temporarily.
Common points include machine status (Run/Idle/Down), cycle time, part counts, alarm codes, feed rate, and spindle load.
MTConnect is an open standard protocol (XML-based) that allows manufacturing equipment to exchange data with software applications.
Modern CNCs use Ethernet (FOCAS/MTConnect); older CNCs use RS-232 serial ports or hardware I/O kits wired to relays.
A heartbeat signal is a periodic electronic pulse sent by a machine to confirm it is powered on and connected to the network.
Yes. You can use external sensors (vibration, current, or optical) to detect cycles on machines without Programmable Logic Controllers.
Human operators often estimate timestamps, round up numbers, or forget to log short downtime events, leading to data accuracy errors of 10-20%.
Nearly any industrial asset, including CNCs (mills/lathes), injection molders, stamping presses, manual saws, and packaging lines.
OEE (Overall Equipment Effectiveness) is a metric calculating the percentage of manufacturing time that is truly productive (Availability × Performance × Quality).
Machine monitoring tracks the asset (uptime, speed, health), while production monitoring tracks the job (part counts, work orders, scheduling).
It connects hardware (sensors, PLCs, or IO adapters) to your manufacturing equipment to capture signals, which are sent to a software dashboard for analysis.
Machine monitoring is the automated process of tracking machine status (running, idle, down) and production data in real-time to identify inefficiencies.
No. Monitoring systems can connect to "legacy" analog machines using digital I/O adapters or current transducers (CT clamps).
.png)
See how Caddis can provide real-time machine insights and proven playbooks to improve your plant operations on Day 1.
Book a Demo
.png){kind=small}