Bitte decken Sie die schraffierte Fläche mit einem Bild ab.
Please cover the shaded area with a picture.
(24,4 x 18,0 cm)

Non-Calibrated High Pressure Sensor

www.continental-corporation.com

Sensors & Actuators Plant Trutnov

Basic information:
Affected project - HPS:
Affected amount of parts:

All analog and digital Projects
110 pcs of HPS total OEM1 World Wide / 212 Sensors out of TRU Plant

Known behavior of parts which are not calibrated:
Output of the sensor signal could be out of specification (see pictures below on this page)
Effect of the wrong sensor´s output could be:
 Lower or higher power of the engine (depends on offset from specification limits).
Acc. to IP results, there is very low risk of overpressure of the system.
 Influence of engine emission (environmental issue).
 There will be not the error message coming from sensor about wrong pressure in rail.

Space for Sender Information
Confidential

7 August 2018
Author, © Continental AG

2

Problem description
Parts were not calibrated because of mishandling with parts before calibration at Continental S&A Trutnov Location before being
provided to Continental Newport News, Pisa and Changchun for assembly into SGE FCA as described herein. This was not able to be
detected after FCA assembly as FCA plants check for electrical function and lack of calibration is detected when parts are pressurized in
system.
Parts were switched in position in tray used for calibration which caused that one part has data belongs to second part. It means first part
was scrapped on code verification and the second part sent to calibration few hours later has in traceability data from previous part and
was not calibrated (because traceability has data, that the part was already calibrated).
The manual handling with part was done because of project fixtures change in hot temperature chamber (we have to cool down the
chamber, change the fixtures, heat up the chamber) but previous processes can not be stopped. Therefore we have to move parts to
small stock in front of calibration. Parts are stored in KLTs.
When calibration is running, we have to move the parts back to trays – physically and also in traceability. This movement is done by
operator.

What happened at Trutnov S&A…
1)

One (first) part was scanned into tray 1, physically moved back to blister because of any issue occures during scanning and into the
tray placed different part – this is the operator mistake.

2)

Different part was sent through the line and was scrapped on DMX reader because system found that this part is not matching the
position in the tray.

3)

After few hours the first part was scanned again into the tray (this is the operator mistake) and sent through the line but:
a)
Part has information about calibration from the different part – therefore our first part was not calibrated again
b)
Part has information about Bin test A from the different part – therefore our first part was not Bin A tested.
c)
DMX reader check if our part match the position in tray and it was correct.

4)

Part was sent to next operation (100 N automatic test) because system has correct information about

5)

Part was correctly processed through all station next stations, packed into the customer blisters and sent to customers.

Space for Sender Information
Confidential

previous processes.

7 August 2018
Author, © Continental AG

3

Visualization of root cause – basic overview
Standard flow without failure:

Flow caused the failure:
Connector welding

Connector welding

Small stock before calibration
because of some break down
+ manual movement

Calibration
Bin test A

Calibration
DMC check
Bin test A + Eletrical
test

Automatic 100N test

DMC check

Bin B electrical test

Automatic 100N test

Final laser marking

Bin B electrical test
Final laser marking
Space for Sender Information
Confidential

7 August 2018
Author, © Continental AG

4

Visualization of failed part flow
123456789

Connector welding
Small stock before calibration
because of some break down
+ manual movement

Parts were placed
into the blister
without traceability.
Small stock not needed and parts
scanned into tha tray where
operator did mistake (mishandling).
Traceability flow.

Calibration
Bin test A + Eletrical
test

1
0

1

Physical flow of mixed parts.

123456789

1
0

123456789

1
0

Physical flow of part found at
customer.

?b c de f gh i j

123456789

1
0

?b c de f gh i j

123456789

1
0

Part 1 has data from part ? That
calibration was done correctly.

Part 1 has data from part ? That
Bin test was done correctly.

DMC check

123456789

1
0

?Part was scrapped on DMC

Automatic 100N test

123456789

1
0

check because the position
in tray was not match the
part ID.

Bin B electrical test

123456789

1
0

123456789

1
0

Final laser marking

123456789

1
0

123456789

1
0

Space for Sender Information
Confidential

123456789

1
0

123456789

1
0

7 August 2018
Author, © Continental AG

5

Failure reproduction – the same condition as before all corrective actions implemented.
Steps of failure reproduction:
1. Scan 3 parts into the tray and replace them with another parts.
a) Parts sent through calibration – OK result.
b) Parts sent through Bin test A – OK result.
c) Parts sent to DMC reader – NOK result – ID of parts doesn´t match the position in tray.
d) Parts were automatically scrapped.
2. Scan 3 original parts again into the tray.
a) Parts correctlly scanned into the tray
b) Parts sent through calibration – calibration not done because traceability has information about OK result from
other/swapped part.
c) Parts sent through Bin test A – Bin test A not done because traceability has information about OK result from
other/swapped part.
d) Parts sent to DMC reader – OK result – the data in traceability were „refreshed“ and the status of part was OK.
e) Next proces steps OK without any issue.
Failure reproduction was successful and confirm our root cause. We are able to switch ON and OFF the failure.
Traceability information – part received from Customer

Traceability of part used for failure reproduction

The same failure message in traceability system found – „checkin_fail; repeat“

Space for Sender Information
Confidential

7 August 2018
Author, © Continental AG

6