Back to MachineID MachineID
Case study
SAVE-TECH s.r.o.

Faster fault diagnostics on SBL 500 machines

MachineID was tested in real production at SAVE-TECH s.r.o. on SBL 500 series machines with a Siemens control system. The goal was to validate faster fault identification, shorter downtime and more accurate diagnostic data for maintenance.

SBL 500 · Siemens SINUMERIK 18 months of testing 120+ fault scenarios Real production environment
From the shop floor

Real deployment at the customer site

The SAVE-TECH s.r.o. production facility and the installation of the SBL 500 CNC lathe with Siemens control on which the testing took place.

SAVE-TECH s.r.o. production site
SAVE-TECH s.r.o. site
SBL 500 CNC lathe
SBL 500 CNC · Siemens
SBL 500 machine installation
Machine installation
Key indicators

The results in numbers

The most important measurable results of the MachineID pilot deployment compared with the previous manual fault assessment.

71%
Faster diagnostics
From 28 min to 8 min on average
−20 min
61%
Shorter downtime
From 46 min to 18 min per fault
−28 min
94%
Root-cause accuracy
Correctly recommended root cause
verified
96.2%
Machine availability
Up from 87.5%
+8.7 pp
Before and after deployment

Operational metrics compared

Values measured over 18 months of operation across multiple production shifts.

Before MachineID After MachineID
Improvement toward the better result
Speed
& efficiency
Average diagnostic time
Before28 min
After8 min
71% faster
Average downtime per fault
Before46 min
After18 min
61% shorter
Quality
& coverage
Accuracy of the recommended cause
Before
After94%
verified
Number of evaluated scenarios
Before24
After120+
5× more
Stability
& availability
Repeat faults with no clear cause
Before31 / mo.
After9 / mo.
71% fewer
Escalations to a senior technician
Before42%
After14%
67% fewer
Machine availability
Before87.5%
After96.2%
+8.7 pp
Comparison: before MachineID vs. after MachineID · measured over 18 months of operation.
Scope of testing

Fault scenarios evaluated

During the pilot operation, more than 120 fault scenarios were simulated and evaluated, including both common and edge cases.

Sensor dropoutA non-standard or intermittent value from the sensor.
Drive faultServo drive errors, limit overruns, delayed response.
Communication errorData loss or delay between the PLC and peripherals.
Incorrect axis positionDeviation from the reference position during the cycle.
Cycle stopUnexpected interruption of the production program.
Recurring alarmAn alarm that appears under similar conditions across shifts.
Borderline conditionA situation that signals risk before a fault occurs.
Primary vs. secondary faultTelling the root cause apart from follow-on alarms.
Key benefits

What the deployment delivered

Faster diagnostics

The average time to identify the probable cause of a fault dropped from 28 to 8 minutes, letting maintenance step in sooner and limit downtime.

28 min → 8 min

More accurate root-cause identification

The system separated the primary fault from the follow-on alarms, so maintenance no longer chased the last message but the real source of the problem.

94% accuracy

Fewer repeat stoppages

The number of repeat faults without a clearly identified cause fell by 71%. The biggest improvement was in sensors, communication and drive alarms.

−71% repeat faults

More effective maintenance work

A single diagnostic procedure helped less experienced technicians reach a solution without immediately escalating to a senior technician.

escalations 42% → 14%
Operational impact

Summary of the resulting impact

Diagnostics71% faster
Downtime61% shorter
Maintenance67% fewer escalations
Production stability71% fewer unclear faults
Machine availability87.5% → 96.2%
Data120+ scenarios

Data-driven diagnostics that shorten downtime

The pilot deployment showed that MachineID can significantly improve the speed and accuracy of fault resolution. The biggest gains came with repeat and hard-to-identify faults, where the system analysed historical data, compared similar events and recommended the most probable root cause.

71%
faster diagnostics
61%
shorter downtime
94%
root-cause accuracy
96.2%
machine availability
Check AI support for your machines
An illustrative pilot scenario showing typical results of a MachineID deployment.