Validation & Verification Study for Metal Detection CCP - AG - LinkedIn profile.pdf

This project contains an example of a
validation and verification study for 3
models of metal detectors posed as an
alternative to replace current aging
models in current food manufacturing
lines.
This is a Validation & Verification for BRC
Global Standards Course Assignment,
MAR 2022
VALIDATION &
VERIFICATION
STUDY OF
METAL
DETECTORS
FOR BRC GLOBAL STANDARDS
Adrian Garcia
DISCLAIMER (please read first):
Neither the data regarding MD performance nor the conclusions of the study are real and have been created as a showcase of what it is expected to appear in best
practice verification studies. A metal detector from any brand can be suitable for one application and not for another, hence the need to do a validation study of
the unit first. There is no ill will meant for any of the brands appearing in this project. However, if you wished to see your brand and/or MD model removed, please
contact me by email at adoriangp@gmail.com. Thank you for your comprehension and please accept my apologies for the inconvenience.

VALIDATION & VERIFICATION STUDY FOR METAL DETECTION CCP
V&V COURSE ASSIGNMENT COMPLETED BY:
ADRIAN GARCIA (SENIOR QA)
Page 1 of 27 Prepared by: AG // Reviewed by: AD // Authorised By: LA Technical Dept.
VALIDATION & VERIFICATION STUDY FOR: NEW metal detector for new packing line (CCP#3 in process).
Contents
1 SCOPE OF THE VALIDATION STUDY...........................................................................................................................2
1.1 Product..............................................................................................................................................................2
1.2 Process ..............................................................................................................................................................2
1.3 Activity ..............................................................................................................................................................3
1.4 Desired Outcome ..............................................................................................................................................3
2 REVIEW OF EXISTING DATA.......................................................................................................................................3
2.1 Historical Data...................................................................................................................................................4
2.2 Data from Codes of Practice & Industry Guidelines .........................................................................................4
2.3 Data from Equipment Suppliers........................................................................................................................4
3 DESIGNING THE VALIDATION STUDY ........................................................................................................................4
3.1 Considerations for this Validation Study...........................................................................................................6
3.2 Design of the Validation Study..........................................................................................................................6
3.2.1 Sensitivity and consistency of performance test ......................................................................................6
3.2.2 Ease of use test .........................................................................................................................................7
4 DECISION CRITERIA (PASS / FAIL CONDITIONS) ........................................................................................................7
5 APPROVAL OF THE VALIDATION STUDY DESIGN & PARAMETERS............................................................................7
6 PERFORMING THE VALIDATION TRIALS....................................................................................................................8
7 GATHERING & COLLATION OF DATA, DOCUMENTATION AND ANALYSIS OF RESULTS............................................8
7.1 Analysis of the Results of the Validation Tests ...............................................................................................21
7.2 Metal Detector Model Benchmarking - Calculation of Overall Probability of Detection (POD %).................21
7.3 Metal Detector Model Benchmarking - Calculation of Overall False Reject Rate (FRR %).............................23
7.4 Overall Summary of the Validation Trials .......................................................................................................23
8 REVIEW & CONCLUSIONS (OUTPUTS).....................................................................................................................24
8.1 Output.............................................................................................................................................................24
8.2 Follow-up Actions to Carry Out as a Result of this Validation Process...........................................................25
8.3 Re-validation Requirements ...........................................................................................................................25
9 APPROVAL OF THE VALIDATION STUDY OUTPUT...................................................................................................26
10 METHOD USED FOR VERIFICATION.....................................................................................................................27
11 FREQUENCY OF VERIFICATION............................................................................................................................27
12 CRITICAL LIMIT FOR VERIFICATION.....................................................................................................................27
13 PASS / FAIL CRITERIA FOR VERIFICATION............................................................................................................27

VALIDATION STUDY – NEW METAL DETECTORS FOR
METAL DETECTION CCP#3
1 SCOPE OF THE VALIDATION STUDY
Validation of new metal detector options for CCP3 for the new upcoming packing line. In more detail:
1.1 Product
Raw material is green coffee beans with a maximum moisture of 12.5%; finished product is roasted coffee beans
and/or roasted & ground coffee products packed in a protective atmosphere inside thermally-sealed metallised
PET/PP film laminate. Finished product is ambient stable and mostly dry with less than 4% moisture and very low Aw
of 0.2.
The metallised packaging film used allows for hygienic, easy and consistent thermal seal of the product packs ensuring
coffee’s organoleptic properties remain unchanged during the guaranteed shelf-life. However, because of the
metallised nature of the packaging film used, it becomes challenging ensuring the metal detection CCP3 step rejects
consistently only packs with physical metal contaminants and not false rejects.
1.2 Process
Process is food manufacturing of roasted coffee beans and/or roasted & ground coffee products packed in a protective
atmosphere inside thermally-sealed metallised PET/PP film laminate at speeds between min. 24 and max. 36 packs
per minute as of current. The speed of the process largely depends on the speed of the roasting and grinding processes
to generate WiP stock to pack: at this moment the process can pack much faster than the roasting capacity. This will
be corrected in mid Q3 2022 with the addition of a new roaster, which could lead to a revision of the packing speeds
up to an est. avg. of 48 packs per minute, hence dramatically increasing site’s output and ability to cope with customer
demands in terms of volume without having to turn to outsourced 3rd
party roasting & packing.
In the site there are established a number of Pre-requisite Programs (PRP) to cover the essential groundwork of food
safety such as cleaning schedules, pest control, control of customer complaints, glass audits, planned preventative
maintenance, personnel training program, supplier approval process, sample testing program, traceability and recall
procedures, and others.
In addition, there have been identified and validated 3 main CCP in the process:
▪ CCP1: verification via visual mould checks at raw material (green coffee) intake to prevent risk of presence of
chemical contamination (mycotoxins such as Ochratoxin) in the product.
▪ CCP2: verification via destoner pressure checks after roasting stage to prevent risk of presence of physical
contaminants in the product (stones, pieces of metal, pieces of glass, wood, plastic, etc.) denser than a roasted
coffee bean.
▪ CCP3: verification via rejection checks of product packs with established sizes of Ferrous / Non-Ferrous /
Stainless Steel metal contaminants test-sticks inside at the metal detection stage in process. Frequency: at the
beginning of the batch run, every hour during the batch run and at the end of the batch run. Rejection of the
metal packs is tested consecutively and alternatively in between normal product packs (memory test). Failsafe
rejection systems are tested at the start and end of each shift.
The metallised nature of the packaging used poses a challenge for the metal detection step, as the device must be
able to screen out contaminants INSIDE the pack whilst not rejecting normal packs. A great degree of careful
calibration of the metal detector sensitivity settings is required to ensure smooth operation, reducing false positives.

False positive rejections, if not controlled, can easily reduce production yield and efficiency, increasing labour costs,
due to the additional manual handling involved (removal of packs, stripping of packs, inline rework of pack contents).
New non-metallised and recyclable types of packaging are under consideration at this point with several trials already
carried out for feasibility of integration in the process (e.g., the thermal sealing properties must be comparable to
metallised film so the film can correctly and consistently seal with our current systems without issues than can derive
in increased downtime). These types of packaging should allow for increased metal detection sensitivity and reduced
false positives once put into place at the end of Q2 2022.
1.3 Activity
Site has both retail customers and food service customers. Space available in the packing hall #2 is quite tight and the
packing lines are close to each other. The vibrations generated by the constant human movement and the air blowing
systems in place are known for occasionally generating interference with the performance of both inline
checkweighers and metal detectors. Unfortunately, until the refurbishment works carried in packing hall #1 are
actually finished by end of Q2 2022 (removal of asbestos roofing, reworking of floors, remaking of walls), the
machinery in place will have to remain close to each other and additional calibrations will have to be done per day to
maintain sensible systems in good working condition
1.4 Desired Outcome
Besides increased sensitivity, any new metal detection devices must be able to better distinguish metal-detectable
physical contaminants in the packs compared to current aging systems which are at the end of their lifespan (10 years).
This should help to drastically reduce downtime due to stripping of false-reject packs as well as packaging waste.
Even though the initial plan is to purchase only 1 metal detector (MD) for the new packing line, there is CAPEX assigned
for the replacement of all current MD units for new models. Before committing to such large expense, business
management have requested a validation of the 3 best MD options in the market for our product type and packing
capacity. In light of taking the best commercial decision, which will also have an impact not just in terms of food safety
but also in process output and B2B relationships with customers with audit programs on site, the expectation is:
a) Providing strong, data-based arguments in favour of one and best solution in terms of technical capabilities
and consistent performance in a dusty, hectic environment where metallised packaging will still be the main
type of film in use for the next 3 years minimum.
b) Evaluating the capability of the units to network and seamlessly integrate into our systems as well as reviewing
the software solutions provided (must be on par or better than our current systems).
c) Exploring suppliers’ ability to provide the MD unit(s) timely and cost-effectively once the decision is taken, as
well as providing a sound service & maintenance program at the best possible price.
All MD units analysed in the validation study must be new models; once the decision is taken, the first finished unit
must be delivered before end of Q1 2022.
2 REVIEW OF EXISTING DATA
A fire destroyed factory’s warehouse in October 2018. As a consequence, many of the physical records in surrounding
areas were lost either directly to the fire or due to water damage during the fire quenching works.
Unfortunately, because of this the current Technical team has been unable to locate the original validation study for
the current MD that established the approved physical contaminant test sizes used1
, as follows:
1
It is known that the test-piece contaminants are large in size compared to most common food industry standards. This is due to the metallised film used having
an impact on MD differential sensitivity between packaging and contaminant.

Production Line Pack Size
TESTSTICKS required for Metallised Film
Ferrous Non-Ferrous Stainless Steel 316
Sandiacre A/B, Rovema 1, 3, 4 & 5 14g to 499g 3.0mm 4.0mm 8.0mm
Sandiacre A/B, Rovema 1, 3, 4 & 5 500g to 1Kg 4.0mm 5.0mm 8.0mm
2.1 Historical Data
Current MD units, have been working for the last 10 years with no major issues to mention. The main concerns
regarding performance of these units are related to how critical the sensitivity setup is related to the product pack
size, and that the hectic environment can lead to bad performance and constant rejection of false positives (e.g.,
vibrations from fans, coffee beans stuck in the MD infeed belt during start-up, etc.). The QCs in the shop floor and
senior operators are trained to deal with these instances.
Another known issue is that even though the units come with password protection for access to the setup menu, it is
not a very robust system as non-senior operators are able to modify the sensitivity settings which is causing issues.
Only senior operators and the QCs / Technical team are trained to do adjustments with proper understanding.
A recent firmware upgrade for the current units greatly increased sustained performance with less false rejection,
reducing downtime overall. Unfortunately, it was found during that firmware installation process that one MD head is
partially damaged, and another MD head is also decaying in sensitivity due to age. These problems are temporarily
resolved via in-depth tweaking of the sensitivity settings, but it is expected that these units will completely fail at some
point before Q1 or Q2 2023, and in any case their performance will be twitchy at best until they fail. Replacing a MD
head normally makes around 50% of the cost of a new MD unit, and because the current units are old the manufacturer
has informed that no fresh or partially used replacements are readily available, hence any parts would have to be
manufactured from scratch at great cost. This situation basically compels the business to find valid alternatives fast.
Review of non-conformities logged in the last 3 years does not show any entries for metal detection failures leading
to quarantined stock. QA reports show good performance overall with mostly pass for metal checks.
2.2 Data from Codes of Practice & Industry Guidelines
Campden BRI’s “Guideline G5/2 Foreign bodies in foods - guidelines for their prevention, control and detection” was
used to identify and exploit the latest technology for the prevention, control and detection of foreign bodies in foods,
and understand what are the capabilities we need to demand from the manufacturer.
2.3 Data from Equipment Suppliers
Technical guides from manufacturers such as Mettler Toledo, Minebea, Loma, CEIA and others were also reviewed to
understand state of current technology and what is the best type of solution for our site & product demands (dry,
metallised film), as well as best practice to validate the system. Mettler Toledo provided a very good model for
validation studies of metal detectors.
The Technical Team has contacted directly a number of manufacturers to review the catalogue of fit-for-purpose
systems they can provide as well as the servicing scheme. These manufacturers were supplied with a number of
samples of the main product sizes (1kg pillow pack, 500g pillow pack, 227g stabilo pack, 50g sachet) to conduct their
own pre-validation trial testing, in laboratory conditions. As a result, each manufacturer suggested a model of metal
detector deemed fit-for-purpose for our process that best caters to our main product sizes, packaging-type used,
environment, and packing output demands.
3 DESIGNING THE VALIDATION STUDY
The following requirements have been considered initially when considering metal detectors to add to our systems:

• Product type (dry, granular and/or powder-like)
• Product sizes (1kg pillow pack, 500g pillow pack, 227g stabilo pack, 50g sachet)
• Product direct-contact packaging (metallised PET/PP film laminate in all cases)
• Process environment (dust, vibration)
• Packing output demands (up to 48 packs/min)
• Metal detector’s feature set including networking / remote control / statistics / data protection
• Settings and product programs protection
• Warranty provided by the manufacturer
• Servicing scheme from the manufacturer / supplier
• Pricing of the unit + yearly servicing (x2 per year)
The Technical team discussed the different MD options suggested by the manufacturers with the Operations team.
Upon this discussion, it was agreed to do an on-site trial and validation of the 3 most promising MD models from 3
long-established makes to find the best solution for the new line. If the model performed as expected, it will be
deployed as a substitute of all the current, aging models for the rest of the lines at a later date.
The proposed models for the validation trial and their suitability to our operation are summarised in the table below:
Model
Mettler-Toledo M33R PlusLine MD Loma Systems IQ4 Metal Detector CEIA THS/MS21 Metal Detector
Make Mettler-Toledo Group Loma Systems CEIA S.p.A.
Housing material Stainless Steel 316L Stainless Steel 316L Stainless Steel 316L
Size of unit (mm) with
volume footprint
**
700 x 405 x 285 (80.8 m3)
*
715 x 400 x 295 (84.4 m3)
***
650 x 395 x 290 (74.5 m3)
What is included
***
MD unit, infeed belt, reject box, sensors
***
***
Dust / water /
vibration protection
***
Heavy duty IP69k
***
Heavy duty IP69k
***
Heavy duty IP69k
Sensitivity to
metallised film
***
Ultrahigh
***
Ultrahigh
***
Ultrahigh
Networking
**
Ethernet
**
Ethernet
***
Ethernet + Wi-Fi 802.11b/g
Data security,
integrity, traceability
***
Password protected, batch log, remote
management software
***
management software
***
management software
Compliance
***
BRCGS, UK/EU Retailers, FDA
**
BRCGS, UK/EU Retailers
**
BRCGS, UK/EU Retailers
Warranty
**
10 years from date of commission
**
*
Servicing type
***
Direct from manufacturer
***
Direct from manufacturer
**
Via 3rd
party partner in UK
Price of unit + yearly
service
*
Most expensive, est. RoI 4.5 years
**
Average, est. RoI 4 years
***
Most affordable, est. RoI 3.5 years
Return of investment (RoI) for these units estimated via sum of pack reject quantities recorded by Production KPI in
the period from end of Q3 2020 to end of Q3 2021, and then calculation of manpower hours required for stripping /
handling of these rejects (avg. of 5 seconds per pack) and an estimation of quantities of film wasted by pack size.
Finally, an estimation of the power consumed in the packing hall during those man hours has also been considered.

3.1 Considerations for this Validation Study
The most important factor for good performance of a metal detector is the dimensions of the product as this will
determine the aperture size of the unit. The sensitivity of the unit is the lowest at the geometrical centre of the
aperture, which is what will be challenged during the validation study.
Once the manufacturers we selected finish their pack trials, they will supply us with trial units with an aperture size fit
for the largest 1kg pack as these will be the units used across all packing lines. The reason is to reduce servicing and
maintenance costs as well as to allow the business flexibility to move the units between lines if needed without having
to do any special adjustments. The unit(s) finally purchased will be manufactured on order and not off-the-shelf.
Other factors affecting sensitivity are:
• Product type
• Metal contaminant type
• Orientation of incoming product packs
The goal of this validation as not just suitability of application, but to find out what are the smallest metal contaminant
test pieces these units can consistently reject to test our products. In addition, for this validation we will also consider
the performance under real life conditions with very few false rejects, ease of product setup and the failsafe process.
3.2 Design of the Validation Study
Considering these factors and the end goal of being able to replace all units in the packing hall in the longer run,
Technical, Production and Engineering teams agreed to carry out 2 validation test types on each unit.
3.2.1 Sensitivity and consistency of performance test
a. Metal contaminant test sticks of Ferrous, Non-Ferrous and Stainless Steel will be tried in sample packs for each
product size considered (1kg pillow pack, 500g pillow pack, 227g stabilo pack, 50g sachet).
b. These metal test packs will be tried for 3 consecutive batch runs for each metal contaminant size to test
consistency of performance.
c. Batch runs must be similar in volume so results are comparable: it was agreed to manufacture and pass
through the metal detector exactly 30 packs per batch run so results are statistically significant enough to
carry out the validation process quickly in order not to affect production capacity.
d. The metal packs will be sequentially tested at the start of a test batch run, after running 15 packs and at the
end of the run. Care must be placed in the way the packs are run through the metal detector’s aperture: their
orientation must be as straight as possible, and as aligned with its centre as possible to truly challenge the
unit. Spacing between the packs must be similar to the normal distance the products travel on the belt. Packing
speed to be set at 36 packs per minute (current standard). It was decided not to include the alternative metal
– non-metal memory test as part of the validation as this would very significantly increase the trial test time
and would require longer runs and more downtime.
e. We will start testing metal contaminants of current site standard sizes, and then repeat the test with
increasingly smaller contaminants (0.5mm at a time) until the unit fails to reject the contaminant once. The
test will continue with the other contaminant types until all 3 types are failed at least once. This moment will
mark the end of the trial for that particular pack size. An additional trial of 3 30-pack runs will be carried out
with the smallest contaminant pieces identified for each pack size to confirm consistency of performance.
f. This process is to be repeated until all pack sizes are tested in each unit. The quantity of packs run, metal test
packs rejected and any false positive rejected packs during the trials must be quantified to calculate the
probability of detection (POD %) and the false reject rate (FRR %) for each metal detector.

3.2.2 Ease of use test
a. Product to be run must be setup by Technical and selected by the Packing Operator, comments on how easy
and quick is to carry out these tasks will be recorded for review and feedback to the manufacturer.
b. Failsafe procedure will be also tested at the beginning of every test for every unit, and any relevant comments
are to be recorded for review and feedback to the manufacturer.
The information collected will be tabulated per unit. To aid in the final decision of this validation process based on
benchmarking, POD and FRR scores are to be calculated as per the following formulas:
Probability of Detection (POD %) False Reject Rate (FRR %)
𝑃𝑂𝐷 % =
𝑅𝑒𝑗𝑒𝑐𝑡𝑒𝑑 𝑇𝑒𝑠𝑡 𝑆𝑎𝑚𝑝𝑙𝑒𝑠
𝑇𝑜𝑡𝑎𝑙 𝑆𝑎𝑚𝑝𝑙𝑒𝑠 𝐼𝑛𝑠𝑝𝑒𝑐𝑡𝑒𝑑
𝑥 100 𝐹𝑅𝑅 % =
𝑇𝑜𝑡𝑎𝑙 𝑅𝑒𝑗𝑒𝑐𝑡 − 𝐶𝑜𝑟𝑟𝑒𝑐𝑡 𝑅𝑒𝑗𝑒𝑐𝑡
𝑇𝑜𝑡𝑎𝑙 𝑆𝑎𝑚𝑝𝑙𝑒𝑠 𝐼𝑛𝑠𝑝𝑒𝑐𝑡𝑒𝑑
𝑥 100
4 DECISION CRITERIA (PASS / FAIL CONDITIONS)
With regards to the POD % calculation during the sensitivity and consistency of performance test:
• PASS: The test pack with the metal contaminant is detected & rejected 100% of the time.
• FAIL: The test pack with the metal contaminant is either not detected or not rejected 100% of the time.
With regards to the FRR % calculation:
• PASS: Less than or equal to 0.5% (5 for every thousand packs).
• FAIL: More than 0.5%.
With regards to the ease of use:
• PASS: 15 seconds or less to unlock, find and select correct product type in the unit during initial setup. Failsafe
can be carried out in 2 minutes (120 seconds) or less.
• FAIL: More than 15 seconds to unlock, find and select the product to run in the unit. Failsafe procedure
requires more than 2 minutes.
5 APPROVAL OF THE VALIDATION STUDY DESIGN & PARAMETERS
The above validation design and scoring parameters have been reviewed and approved by the Heads of each involved
department, as well as reviewed and approved by the Site General Manager:
VALIDATION TRIAL APPROVAL – 11/10/2021
Position Signature
Technical Manager Signature of Site’s Technical Manager
Production Manager Signature of Site’s Production Manager
Engineering Manager Signature of Site’s Engineering Manager
Site General Manager Signature of Site’s General Manager

6 PERFORMING THE VALIDATION TRIALS
After the approval of the validation study and its scoring metrics it was discussed with the 3 chosen manufacturers
availability of test units. Because of the initial trials done by the manufacturers with representative pack samples of
the different product sizes to test, we could quickly confirm that all of the metal detector test units would be on site
by the 1st
week of January 2022 in order not to disrupt production during the Christmas season.
The validation trials were hence scheduled for w/c 10/01/2022 to last 3 days. Supply Chain created all of the
corresponding batch runs in the systems, and allocated a space in the warehouse for all of the resulting samples
generated. Once the validation is completed, the packs will be passed through the current metal detectors and if
passed, added to the stock. After discussion with Production & Engineering, it was decided to test roast & ground
coffee for the smaller pack sizes (50g & 227g) and coffee beans for the larger pack sizes (500g & 1000g) as most of the
volumes for each product size corresponds to those product types, hence it would be a more realistic representation
of actual production.
Along that week, the Senior QA responsible for the management of systems on site together with a Senior Operator
and the support of the QC on shift will team up to carry out the different tests as outlined in the validation design
section. Engineering Manager will support the process as required. Each trial metal detector will be set on a line every
day, then swap around during the nightshift until all of them are tried in each product size. A Field Engineer from
Mettler-Toledo will help with the trials for their unit during the days it is tried with the main product types (1Kg, 227g);
the individual will also lend a whole collection of test stick sizes to support the process. The other 2 manufacturers
were unable to provide such support.
It is expected that the task will take at least shift for the individuals per each trial day until the validation is completed
including all of the scheduled tests.
Senior QA will document all of the results obtained so they can be plotted on a table for later review.
7 GATHERING & COLLATION OF DATA, DOCUMENTATION AND ANALYSIS OF RESULTS
The validation trials of the 3 metal detector models took place during days 10, 11 & 12 of January 2022 as scheduled.
A total of 339 short 30-pack runs over the 3-day period were carried out. At the set packing speed of 36 packs/min,
each run took about a minute to complete, plus the time required for the actual metal detection tests and preparation
of the test packs (an average of 2 minutes was required per test run). The start-up failsafe and product setup tests
where factored in as well.
The results of these trial runs for each pack size and metal detector were plotted onto the following tables with the
sequence of the different metal test sizes tested.

Product Type: Roast & Ground Coffee (Grind Setting A)
Pack Size: 50g
Model Tested:
Mettler-Toledo M33R PlusLine Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS
Metal Contaminant Size 3.0 4.0 8.0 2.5 3.5 7.5 2.0 3.0 7.0 1.5 2.5 6.5 1.0 2.0 6.0 0.5 1.5 5.5 1.0 2.0 5.0 1.0 2.0 4.5 1.0 2.0 4.0
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run MT-50-1a Trial Run MT-50-2a Trial Run MT-50-3a Trial Run MT-50-4a Trial Run MT-50-5a Trial Run MT-50-6a Trial Run MT-50-7a Trial Run MT-50-8a Trial Run MT-50-9a
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 1/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run MT-50-1b Trial Run MT-50-2b Trial Run MT-50-3b Trial Run MT-50-4b Trial Run MT-50-5b Trial Run MT-50-6b Trial Run MT-50-7b Trial Run MT-50-8b Trial Run MT-50-9b
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run MT-50-1c Trial Run MT-50-2c Trial Run MT-50-3c Trial Run MT-50-4c Trial Run MT-50-5c Trial Run MT-50-6c Trial Run MT-50-7c Trial Run MT-50-8c Trial Run MT-50-9c
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 93% 100% 100% 96%
FRR % per trial (90
packs)
0% 0% 4% 0% 0% 0% 0% 0% 0%
Test Results (Metrics)
Smallest Metal Contaminant Sizes Validated for the Pack
Size
PASS Condition:
Failsafe Test Time (sec.) 112 Original Size
NEW Validated
Size
Rejection of the metal contaminant size at start, mid and end of the test
run.
Product Setup Time
(sec.)
9 Ferrous Fe 3.0 mm 1.0 mm
Overall POD % 98.8% Non-Ferrous NFe 4.0 mm 2.0 mm
Overall FRR % 0.4% Stainless Steel SS 8.0 mm 4.5 mm

Pack Size: 50g
Model Tested:
Loma Systems IQ4 Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Lo-50-1a Trial Run Lo-50-2a Trial Run Lo-50-3a Trial Run Lo-50-4a Trial Run Lo-50-5a Trial Run Lo-50-6a Trial Run Lo-50-7a Trial Run Lo-50-8a Trial Run Lo-50-9a
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Lo-50-1b Trial Run Lo-50-2b Trial Run Lo-50-3b Trial Run Lo-50-4b Trial Run Lo-50-5b Trial Run Lo-50-6b Trial Run Lo-50-7b Trial Run Lo-50-8b Trial Run Lo-50-9b
No
Reject
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Lo-50-1c Trial Run Lo-50-2c Trial Run Lo-50-3c Trial Run Lo-50-4c Trial Run Lo-50-5c Trial Run Lo-50-6c Trial Run Lo-50-7c Trial Run Lo-50-8c Trial Run Lo-50-9c
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 1/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 96% 96% 100% 100% 93%
FRR % per trial (90
packs)
0% 0% 0% 0% 4% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Pack Size: 50g
Model Tested:
CEIA THS/MS21 Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test
Size
Original
Test
Size
Original
Test
Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
-0.5mm
from last
test
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Ce-50-1a Trial Run Ce-50-2a Trial Run Ce-50-3a Trial Run Ce-50-4a Trial Run Ce-50-5a Trial Run Ce-50-6a Trial Run Ce-50-7a Trial Run Ce-50-8a Trial Run Ce-50-9a
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Ce-50-1b Trial Run Ce-50-2b Trial Run Ce-50-3b Trial Run Ce-50-4b Trial Run Ce-50-5b Trial Run Ce-50-6b Trial Run Ce-50-7b Trial Run Ce-50-8b Trial Run Ce-50-9b
No
Reject
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Ce-50-1c Trial Run Ce-50-2c Trial Run Ce-50-3c Trial Run Ce-50-4c Trial Run Ce-50-5c Trial Run Ce-50-6c Trial Run Ce-50-7c Trial Run Ce-50-8c Trial Run Ce-50-9c
No
Reject
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS FAIL
False Rejects 1/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 96% 89% 100% 89%
FRR % per trial (90
packs)
4% 0% 0% 0% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type: Roast & Ground Coffee (Grind Setting B)
Pack Size: 227g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
Same as
last PASS
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run MT-227-1a Trial Run MT-227-2a Trial Run MT-227-3a Trial Run MT-227-4a Trial Run MT-227-5a Trial Run MT-227-6a Trial Run MT-227-7a Trial Run MT-227-8a Trial Run MT-227-9a
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run MT-227-1b Trial Run MT-227-2b Trial Run MT-227-3b Trial Run MT-227-4b Trial Run MT-227-5b Trial Run MT-227-6b Trial Run MT-227-7b Trial Run MT-227-8b Trial Run MT-227-9b
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run MT-227-1c Trial Run MT-227-2c Trial Run MT-227-3c Trial Run MT-227-4c Trial Run MT-227-5c Trial Run MT-227-6c Trial Run MT-227-7c Trial Run MT-227-8c Trial Run MT-227-9c
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 1/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 96% 96% 100% 100% 96%
FRR % per trial (90
packs)
0% 0% 0% 4% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Pack Size: 227g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Lo-227-1a Trial Run Lo-227-2a Trial Run Lo-227-3a Trial Run Lo-227-4a Trial Run Lo-227-5a Trial Run Lo-227-6a Trial Run Lo-227-7a Trial Run Lo-227-8a Trial Run Lo-227-9a
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Lo-227-1b Trial Run Lo-227-2b Trial Run Lo-227-3b Trial Run Lo-227-4b Trial Run Lo-227-5b Trial Run Lo-227-6b Trial Run Lo-227-7b Trial Run Lo-227-8b Trial Run Lo-227-9b
No
Reject
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Lo-227-1c Trial Run Lo-227-2c Trial Run Lo-227-3c Trial Run Lo-227-4c Trial Run Lo-227-5c Trial Run Lo-227-6c Trial Run Lo-227-7c Trial Run Lo-227-8c Trial Run Lo-227-9c
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 1/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 89% 100% 100% 100% 93%
FRR % per trial (90
packs)
0% 0% 0% 0% 4% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Pack Size: 227g
Model Tested:
CEIA THS/MS21 Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test
Size
Original
Test
Size
Original
Test
Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
-0.5mm
from last
test
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
No
Reject
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 93% 93% 100% 89%
FRR % per trial (90
packs)
0% 0% 0% 0% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 500g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS
Metal Contaminant Size 4.0 5.0 8.0 3.5 4.5 7.5 3.0 4.0 7.0 2.5 3.5 6.5 2.0 3.0 6.0 1.5 2.5 5.5 1.0 2.0 5.0 0.5 1.5 4.5
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run MT-500-1a Trial Run MT-500-2a Trial Run MT-500-3a Trial Run MT-500-4a Trial Run MT-500-5a Trial Run MT-500-6a Trial Run MT-500-7a Trial Run MT-500-8a
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects Rejects Rejects
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run MT-500-1b Trial Run MT-500-2b Trial Run MT-500-3b Trial Run MT-500-4b Trial Run MT-500-5b Trial Run MT-500-6b Trial Run MT-500-7b Trial Run MT-500-8b
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run MT-500-1c Trial Run MT-500-2c Trial Run MT-500-3c Trial Run MT-500-4c Trial Run MT-500-5c Trial Run MT-500-6c Trial Run MT-500-7c Trial Run MT-500-8c
No
Reject
Rejects Rejects Rejects Rejects
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 100% 93% 96%
FRR % per trial (90
packs)
0% 0% 0% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 500g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Lo-500-1a Trial Run Lo-500-2a Trial Run Lo-500-3a Trial Run Lo-500-4a Trial Run Lo-500-5a Trial Run Lo-500-6a Trial Run Lo-500-7a Trial Run Lo-500-8a
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Lo-500-1b Trial Run Lo-500-2b Trial Run Lo-500-3b Trial Run Lo-500-4b Trial Run Lo-500-5b Trial Run Lo-500-6b Trial Run Lo-500-7b Trial Run Lo-500-8b
No
Reject
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Lo-500-1c Trial Run Lo-500-2c Trial Run Lo-500-3c Trial Run Lo-500-4c Trial Run Lo-500-5c Trial Run Lo-500-6c Trial Run Lo-500-7c Trial Run Lo-500-8c
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 1/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 100% 89% 96%
FRR % per trial (90
packs)
0% 0% 0% 0% 0% 0% 4% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 500g
Model Tested:
CEIA THS/MS21 Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test
Size
Original
Test
Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
Same as
last
PASS
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
No
Reject
Rejects
No
Reject
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 100% 96% 89% 93%
FRR % per trial (90
packs)
0% 0% 0% 0% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 1000g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS
Metal Contaminant Size 4.0 5.0 8.0 3.5 4.5 7.5 3.0 4.0 7.0 2.5 3.5 6.5 2.0 3.0 6.0 1.5 2.5 5.5 2.0 3.0 5.0
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run MT-1000-1a Trial Run MT-1000-2a Trial Run MT-1000-3a Trial Run MT-1000-4a Trial Run MT-1000-5a Trial Run MT-1000-6a Trial Run MT-1000-7a
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run MT-1000-1b Trial Run MT-1000-2b Trial Run MT-1000-3b Trial Run MT-1000-4b Trial Run MT-1000-5b Trial Run MT-1000-6b Trial Run MT-1000-7b
No
Reject
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL FAIL PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run MT-1000-1c Trial Run MT-1000-2c Trial Run MT-1000-3c Trial Run MT-1000-4c Trial Run MT-1000-5c Trial Run MT-1000-6c Trial Run MT-1000-7c
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS PASS PASS FAIL
False Rejects 0/30 0/30 1/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 85% 93%
FRR % per trial (90
packs)
0% 0% 4% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 1000g
Model Tested:
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS
Metal Contaminant Size 4.0 5.0 8.0 3.5 4.5 7.5 3.0 4.0 7.0 2.5 3.5 6.5 2.0 3.0 6.0 1.5 3.5 5.5 2.0 3.5 5.0
Validation Sequence
Original
Test Size
Original
Test Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
Same as
last PASS
-0.5mm
from last
test
Same as
last PASS
Same as
last PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Lo-1000-1a Trial Run Lo-1000-2a Trial Run Lo-1000-3a Trial Run Lo-1000-4a Trial Run Lo-1000-5a Trial Run Lo-1000-6a Trial Run Lo-1000-7a
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Lo-1000-1b Trial Run Lo-1000-2b Trial Run Lo-1000-3b Trial Run Lo-1000-4b Trial Run Lo-1000-5b Trial Run Lo-1000-6b Trial Run Lo-1000-7b
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Lo-1000-1c Trial Run Lo-1000-2c Trial Run Lo-1000-3c Trial Run Lo-1000-4c Trial Run Lo-1000-5c Trial Run Lo-1000-6c Trial Run Lo-1000-7c
No
Reject
Mid of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL PASS FAIL PASS PASS PASS PASS FAIL
False Rejects 1/30 0/30 0/30 0/30 0/30 1/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 93% 93% 93%
FRR % per trial (90
packs)
4% 0% 0% 0% 0% 4% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

Product Type:
Roasted Coffee
Beans
Pack Size: 1000g
Model Tested:
CEIA THS/MS21 Metal
Detector
Observations
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Last test
size that
fully
PASSED
Metal Contaminant
Type
Validation Sequence
Original
Test
Size
Original
Test
Size
Original
Test Size
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from
last test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from last
test
-0.5mm
from
last test
Same as
last
PASS
Same as
last
PASS
-0.5mm
from
last test
1st 30 normal packs
run
Trial Run Ce-1000-1a Trial Run Ce-1000-2a Trial Run Ce-1000-3a Trial Run Ce-1000-4a Trial Run Ce-1000-5a Trial Run Ce-1000-6a Trial Run Ce-1000-7a Trial Run Ce-1000-8a
No
Reject
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL FAIL PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
2nd 30 normal packs
run
Trial Run Ce-1000-1b Trial Run Ce-1000-2b Trial Run Ce-1000-3b Trial Run Ce-1000-4b Trial Run Ce-1000-5b Trial Run Ce-1000-6b Trial Run Ce-1000-7b Trial Run Ce-1000-8b
No
Reject
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
3rd 30 normal packs
run
Trial Run Ce-1000-1c Trial Run Ce-1000-2c Trial Run Ce-1000-3c Trial Run Ce-1000-4c Trial Run Ce-1000-5c Trial Run Ce-1000-6c Trial Run Ce-1000-7c Trial Run Ce-1000-8c
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
No
Reject
No
Reject
No
Reject
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS FAIL FAIL PASS PASS PASS FAIL
False Rejects 0/30 0/30 0/30 0/30 0/30 0/30 0/30 0/30
POD % per trial (90
packs)
100% 100% 100% 100% 100% 100% 81% 93%
FRR % per trial (90
packs)
0% 0% 0% 0% 0% 0% 0% 0%
Size
PASS Condition:
NEW Validated
Size
run.
Product Setup Time
(sec.)

7.1 Analysis of the Results of the Validation Tests
As a result of all of the tests carried out, the smallest metal contaminant sizes (in mm) fully rejected per each product
size and metal detector model were identified. See the summary in the table below:
Pack Sizes and Types → 50g R&G 227g R&G 500g Beans 1000g Beans
Contaminant Types
→
Fe-
50g
NFe-
50g
SS-
50g
Fe-
227g
NFe-
227g
SS-
227g
Fe-
500g
NFe-
500g
SS-
500g
Fe-
1000
g
NFe-
1000
g
SS-
1000
g
Mettler-Toledo
M33R PlusLine Metal
Detector
1.0 2.0 4.5 1.5 2.0 4.5 1.5 2.5 5.0 2.0 3.0 5.5
Loma Systems IQ4
Metal Detector
1.0 2.5 4.5 1.5 2.5 4.5 1.5 2.5 5.0 2.0 3.5 5.5
CEIA THS/MS21
Metal Detector
1.0 1.5 4.5 1.0 1.5 4.5 1.0 2.5 4.5 1.5 2.5 5.0
In green is shown the absolute smallest size for each metal contaminant type and product size (best), in orange the
next smallest (average), in red the largest piece of the sequence (worst). See page 4 for current size test pieces. The
CEIA unit seems to have the best sensitivity overall followed by the Mettler-Toledo model. Loma’s metal detector
seems to lag behind with regards to sensitivity to the non-ferrous contaminants as the test packs increase in size.
7.2 Metal Detector Model Benchmarking - Calculation of Overall Probability of Detection (POD %)
Once all of the smallest metal contaminant sizes per each pack size were identified, a final sequence of 3 “30-pack”
trial runs was carried out to test consistency of rejection of these pieces. See following the results of these tests.
Model Tested: Mettler-Toledo M33R PlusLine Metal Detector
Pack Type & Size: R&G (A) 50g R&G (C) 227g Beans 500g Beans 1000g
Metal Contaminant
Type
Fe NFe SS Fe NFe SS Fe NFe SS Fe NFe SS
Metal Contaminant Size 1.0 2.0 4.5 1.5 2.0 4.5 1.5 2.5 5.0 2.0 3.0 5.5
1st 30 normal packs
run
Trial Run MT-50-POD-1a Trial Run MT-227-POD-2a Trial Run MT-500-POD-3a Trial Run MT-1000-POD-4a
Start of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
End of Run Test Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects Rejects
Overall Result PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS PASS
2nd 30 normal packs
run
Trial Run MT-50-POD-1b Trial Run MT-227-POD-2b Trial Run MT-500-POD-3b Trial Run MT-1000-POD-4b
3rd 30 normal packs
run
Trial Run MT-50-POD-1c Trial Run MT-227-POD-2c Trial Run MT-500-POD-3c Trial Run MT-1000-POD-4c
POD % per trial (90
packs)
100% 100% 100% 100%
Test Results (Metrics) PASS Condition:
Overall POD % 100.0%
run.

Model Tested: Loma Systems IQ4 Metal Detector
Metal Contaminant
Type
1st 30 normal packs
run
Trial Run Lo-50-POD-1a Trial Run Lo-227-POD-2a Trial Run Lo-500-POD-3a Trial Run Lo-1000-POD-4a
2nd 30 normal packs
run
Trial Run Lo-50-POD-1b Trial Run Lo-227-POD-2b Trial Run Lo-500-POD-3b Trial Run Lo-1000-POD-4b
3rd 30 normal packs
run
Trial Run Lo-50-POD-1c Trial Run Lo-227-POD-2c Trial Run Lo-500-POD-3c Trial Run Lo-1000-POD-4c
POD % per trial (90
packs)
100% 100% 100% 100%
run.
Model Tested: CEIA THS/MS21 Metal Detector
Metal Contaminant
Type
1st 30 normal packs
run
Trial Run Ce-50-POD-1a Trial Run Ce-227-POD-2a Trial Run Ce-500-POD-3a Trial Run Ce-1000-POD-4a
2nd 30 normal packs
run
Trial Run Ce-50-POD-1b Trial Run Ce-227-POD-2b Trial Run Ce-500-POD-3b Trial Run Ce-1000-POD-4b
3rd 30 normal packs
run
Trial Run Ce-50-POD-1c Trial Run Ce-227-POD-2c Trial Run Ce-500-POD-3c Trial Run Ce-1000-POD-4c
POD % per trial (90
packs)
100% 100% 100% 100%
run.

As a result of these final trial runs, see following the summary table with the resulting calculation of the overall
probability of detection (POD %) score obtained for each metal detector units as per the formula outlined in page 7
(higher percentage is better):
Model
Rejected Test
Samples
Total Valid Samples
Inspected
Overall POD %
Mettler-Toledo M33R PlusLine Metal Detector 108 108 100.00%
Loma Systems IQ4 Metal Detector 108 108 100.00%
CEIA THS/MS21 Metal Detector 108 108 100.00%
All of the smallest metal contaminant test sizes identified were consistently rejected during these trials, which
completes their validation. All of the metal detectors obtain a PASS for this factor as all of them consistently rejected
the tests every time.
7.3 Metal Detector Model Benchmarking - Calculation of Overall False Reject Rate (FRR %)
With regards to the False Reject Rate, this figure was obtained during the initial trial test runs taking into account all
of the packs produced. The final validation tests were excluded from this calculation.
As a result, see following summary table with the resulting calculation of the overall false reject rate (FRR %) score
obtained for each metal detector unit as per the formula outlined in page 7 (smaller percentage is better):
Model Total Rejects
Correct Valid
Rejects
Total Samples
Inspected
Overall FRR %
Mettler-Toledo M33R PlusLine Metal Detector 879 876 891 0.34%
Loma Systems IQ4 Metal Detector 877 872 891 0.56%
CEIA THS/MS21 Metal Detector 921 920 945 0.11%
Taking into account the PASS condition we established in page 7 (less than 0.5%), the unit with lowest (best) FRR was
the CEIA one, followed by the Mettler Toledo. Both of them obtained a PASS for this benchmark. The Loma unit
obtained an FRR above the FAIL threshold, hence failing this test.
7.4 Overall Summary of the Validation Trials
See below summary table of the data gathered during the validation process for each MD unit tested:
Model
Detector
Loma Systems IQ4 Metal Detector CEIA THS/MS21 Metal Detector
Make Mettler-Toledo Group Loma Systems CEIA S.p.A.
Total 30-Pack Runs 99 99 105
Total Pack QTY Run 2,970 2,970 3,150
Total Metal Packs
Tested Count
891 891 945
Total False Rejects
Count
3 5 1
Avg. Failsafe Test
Time (s)
107 106 116

Model
Detector
Loma Systems IQ4 Metal Detector CEIA THS/MS21 Metal Detector
Avg. Product Setup
Time (s)
11 11 15
Comments from
Packing Operator /
QC regarding ease-of-
use
▪ Easy to set the product
▪ Interface easy to follow, amazing
stats
▪ Responsive touchscreen
▪ Easy to set a product & create a run
▪ Interface is easy with lots of
information
▪ Fast to turn on, responsive screen
▪ Screen is small, operator must trust
sound more than image
▪ Interface outdated
▪ Product is easy to set
Overall POD %
***
100% (PASS)
***
100% (PASS)
***
100% (PASS)
Overall FRR %
**
0.34% (PASS)
*
0.56% (FAIL)
***
0.11% (PASS)
Sensitivity
**
Very close to best
*
Struggled with Non-Ferrous test pieces
***
Best overall with smallest test pieces
Ease of use & data
security
***
Modern interface, fast to set, easy
remote management, password per user
***
Modern interface, fast to set, easy
remote management, password per user
*
Outdated interface, reliance on sound,
longest to set, easy remote
management, general password
SUITABILITY GRADE
***
Close to best sensitivity, good useability,
best service, best business partner
**
NFe sensitivity lags, best useability, good
service, good business partner
**
Best sensitivity, however useability,
service from manufacturer lags
8 REVIEW & CONCLUSIONS (OUTPUTS)
All data collected during the validation process will be used by the business:
▪ To provide evidence on what are the absolute smallest metal contaminant test pieces the unit can
consistently detect for each pack size.
▪ To help taking a purchase decision for the best and most suitable unit considering also other commercial
factors such as ongoing service costs.
Upon review of the data and the scoring obtained for the different metal detector models the conclusions were as
following
8.1 Output
The metal detector model most suitable for the requirements of the business based on the data from the trials
(sensitivity, consistency of performance) and considering warranty, servicing program, B2B support provided and ease
of use we believe is the Mettler-Toledo M33R PlusLine Metal Detector. This unit will provide a vast improvement in
metal contamination sensitivity overall for all pack sizes compared to the current MD models, which will be welcomed
by both retailers as well as food service customers as the suitability of our current metal detectors has been questioned
during audits. Also, the software provided to set the different pack size programs as well as the remote management
of the unit is best in the industry and should facilitate and accelerate the new product setup process going forward.

8.2 Follow-up Actions to Carry Out as a Result of this Validation Process
In order to introduce the chosen unit seamlessly into our process, the following action plan was agreed between
Technical and Operations:
1. It was decided to keep a similar 2-set of metal test-stick as of current, with smaller test pieces for 227g sized
packs or below, and bigger test pieces for products above 227g up to 1kg. The approved test-stick setup for
the new Mettler Toledo unit(s) is the following (current MD units to keep old test-stick sizes):
2. Technical team to purchase the new sets of metal test-sticks for the new MD.
3. Technical team to update the HACCP manual regarding CCP 3 by adding the above table for the Mettler-Toledo
metal detectors as reference of this validation. The test-stick table for the current units will still need to remain
in the HACCP until all of these units are replaced with the new ones.
4. Technical will communicate these changes in the HACCP manual to Operations Management for their review
and approval.
5. A number of work instructions will need updating by the Technical team:
▪ WI.CO.035 How to Conduct a Metal Detector Quality Check Test: to be updated to v21 to include the table
above as reference for the Mettler-Toledo units.
▪ WI.CO.037 How to Conduct a Failsafe Check: to be updated to v3 to add the new failsafe check flow for
the new MD units.
▪ WI.CO.038b How to Setup the Combi-unit (Checkweigher - Metal Detector): to be updated to v7 to add
the new process for the Mettler-Toledo metal detectors.
6. A number of forms will need updating by the Technical team:
▪ PC014a Missing Metal Test-Sticks Control Sheet: to be updated to v4 to add checks for the new smaller
test-stick sizes.
▪ PC014b Failsafe Daily Checklist: to be updated to v5 to include checklist for the new metal detector.
7. Electronic Q-Sys Qlog QA software to be updated to include the new approved test-stick sizes and a new metal
detector selection screen during batch run setup for the Operator.
8. Technical & Operations will collaborate to train the Packing Operators and the QC on these updated work
instructions and forms.
9. Verification of Failsafe procedure to be carried out by the Packing Operator at the same frequencies as for the
current metal detectors and to be recorded onto updated PC014b Failsafe Daily Checklist v5.
10. Verification of detection and rejection of sample packs with metal contaminant test-sticks inside to happen at
the same frequencies as for the current metal detectors, and to be recorded into the updated electronic Q-
Sys Qlog QA software as usual.
8.3 Re-validation Requirements
The chosen MD will receive a 6-monthly service & calibration regime by the manufacturer where the unit will also be
tested with the approved test-stick sizes to ensure it detects and rejects and the metal detector head remains healthy.
These re-validation procedures will be recorded in the service report.
In the case there is no detection or rejection of one of the metal test-sticks during these calibrations, once the issue is
confirmed to be related to the aging of the detector head, the unit will have to undergo a full re-validation process to
find out which test piece is now the smallest contaminant size (of the same kind that initially failed the test) that can
be consistently detected and rejected. In this case the re-validation will be carried out following the same process

outlined before (3 consecutive 30-pack batch runs with 3 metal detector checks in each run, where the piece is
detected and rejected always).
9 APPROVAL OF THE VALIDATION STUDY OUTPUT
The output and follow-up action list has been reviewed and approved by the Heads of each involved department, as
well as reviewed and approved by the Site General Manager:
VALIDATION TRIAL OUTPUT – 24/01/2022
Position Signature
Technical Manager Signature of Site’s Technical Manager
Production Manager Signature of Site’s Production Manager
Engineering Manager Signature of Site’s Engineering Manager
Site General Manager Signature of Site’s General Manager

VERIFICATION OF EFFECTIVENESS OF NEW METAL
DETECTORS
10 METHOD USED FOR VERIFICATION
Metal detection is a control measure for removal of metallic foreign bodies from the product. In our system, this is
also CCP3 as it is the last step in the process where this control can take place.
The control measure method is checking the metal detector detection and rejection by carrying out a consecutive
and a memory test with 3 metal test packs of the same size as the product being packed at scheduled intervals.
Each of these test packs must be prepared with a metal test-stick inside containing a Ferrous, Non-Ferrous and
Stainless-Steel metal contaminant of previously validated and approved sizes.
11 FREQUENCY OF VERIFICATION
The effectiveness of the metal detection CCP3 is to be verified as per the below scheduled intervals:
▪ At the start of a batch run.
▪ Every hour during the batch run.
▪ And at the end of the batch run.
12 CRITICAL LIMIT FOR VERIFICATION
The limit of detectability depends on pack size.
The approved contaminant test sizes for all current metal detectors on site are:
The approved contaminant test sizes for all Mettler-Toledo M33R PlusLine metal detector(s) to be commissioned are:
13 PASS / FAIL CRITERIA FOR VERIFICATION
See below pass/fail criteria for metal detection verification assessment (check) to identify when the process is not
effectively delivering the required level of operation:
▪ PASS: all metal test packs are detected and rejected in both the consecutive and the alternative (memory)
tests.
▪ FAIL: at least one metal test pack is not detected or not rejected during the metal detector check.

Validation & Verification Study for Metal Detection CCP - AG - LinkedIn profile.pdf

More Related Content

What's hot (20)

Similar to Validation & Verification Study for Metal Detection CCP - AG - LinkedIn profile.pdf (20)

Recently uploaded (20)

Validation & Verification Study for Metal Detection CCP - AG - LinkedIn profile.pdf