Category Archives: Animal health

Ekomilk-AMP system description

PDF: system description

This introductory description (pdf) gives an idea of the key features of the Ekomilk-AMP system. This system includes Ekomilk Horizon, the AMPI Sample Feeder and essential modules of the Cloud-based Animal Monitoring Platform (AMP) which allows to monitor and compare bulk tanks, individual dairy animals and udder quarters. Variation on this system leads for instance to the version for milking robots and versions to integrate together with other equipment.

PDF: system description

Acidosis and ketosis monitor

Thanks to Ekomilk Horizon Unlimited+ and the online ketosis and acidosis monitoring module, for every individual milk sample, the likelihood of ketosis and acidosis is investigated. The AMP notification system can easily be configured in your user account to define thresholds for sending SMS, email, online app notifications (and messages for custom XML/JSON protocols) together with online AMP dashboards per cow, tank and herd.

For an introductory and abstract description of the Ekomilk – AMP system, click on the button:
Ekomilk-AMP system description


This ketosis (hyperketonemia) and acidosis monitoring module uses algorithms based on research by leading universities worldwide complemented by our own investigations.
This module is optimised for specificity, which according to studies and simulations is usually over 80% (field studies still on-going, Q4 2017 – Q1 2018). It is important to regularly (e.g. at least monthly, frequency to be determined) perform individual (per cow, buffalo, other) milk analysis using ICAR samplers, followed by a correct identification of each sample (cow, buffalo, other animal).
The accuracy for this individual ketosis monitoring will increase when more regular measurement data is available for the cow and for the herd. For better accuracy, measure regularly.

Workflow and cost:

This individual identification is greatly facilitated by AMP, thanks to its templating system which allows barcode reader integration and the USB keyboards which can be connected to Ekomilk Horizon.
Monthly cost of this additional AMP module: 8 euro EXW for up to 200 cows.
The cost per test remains the same at approx. €0.03 EXW.

Available from Q1 2018. For more information please contact us via the contact form.

Introduction video of Ekomilk Horizon, AMPI Sample Feeder and common online AMP modules:

Background information for ketosis (hyperketonemia) :
Ketosis is a process whereby the concentration of ketone bodies in the serum (blood) increases significantly. The 3 ketone bodies are BHB, acetoacetate (AcAc) and acetone (Ac).
Also in milk certain increases can be found although the correlation factor between serum and milk concentration is not always high (more info coming).
Ketotic cases indicate an imbalance of the metabolic processes, for instance postpartum (after calving, birth). There is a negative energy balance. Such imbalance can also be caused by insufficient carbohydrate intake. In a ketotic case, ketone bodies will provide energy to the animal (dairy cow or buffalo) instead of the usual glucose as an energy source.

Milk composition: milk fat, protein, total solids content and pH, added water

For accurate milk composition analysis: fat (butterfat), protein, total solids content and pH, conductivity, added water at $0.04/test, Ekomilk Horizon Unlimited together with the automatic AMPI Sample Feeder is the best solution for many users, whether in a traditional milking parlor, for direct connection to a milking robot or for laboratory application. Ekomilk Horizon also performs accurate somatic cell count. Every Ekomilk Horizon Unlimited includes the essential AMP modules. Extra AMP modules and custom design can be added afterwards.

For an introductory and abstract description of the Ekomilk – AMP system, click on the button:
Ekomilk-AMP system description

Please click on the following topics for more info:


Configure here your own Ekomilk-AMP solution

Ekomilk - AMP direct connection to milking robot (simulation)

Milking robots: automatic milk quality and milk composition analysis without Shuttle at $0.04 per test

Milking robots: introduction
Milking robots are growing in popularity on dairy farms worldwide.
Robotic milking, or automatic milking systems (AMS) implies that the cow more or less chooses when she will be milked. With robotic milking, the cow can be milked at any point during the day, also during the night. Changing the milking system in the milking parlor requires a transition period for both the farmer and the dairy cows.
The most famous companies and brands (models) for milking robots are probably these ones:

  • Lely Astronaut, A3/A4,
  • DeLaval VMS (VMS refers to voluntary milking),
  • Boumatic MR-S1
  • Fullwood Merlin,
  • SAC (SA Christensen) Futureline

Feel free to contact us with your brand suggestion.

For an introductory and abstract description of the Ekomilk – AMP system, click on the button:
Ekomilk-AMP system description

Sometimes we still hear the idea that via milking robots, everything will be automated and no attention should be paid to the dairy cows at all. As we heard over and again, such attitude can lead to illnesses spreading in the herd and decreasing milk yields, followed by lower milk revenues, sometimes aggravated by lower milk quality and according penalties in case of high somatic cell counts (SCC; subclinical and clinical mastitis). Longevity will usually suffer as well. Click here for a short introduction to somatic cell count, milk quality and milk yield, mastitis and udder health.

Animal health, milk composition and milk quality, mastitis and udder health
What’s going on with your dairy cows, how likely is it that they will have a good milk yield (production) and milk quality (somatic cells, fat, protein and solid content levels) in near and mid-term (future)?
How likely is it that the cow feeding has to be changed for a specific animal in the herd on the dairy farm? Any metabolic issues? Acidosis or ketosis? How is the protein and fat content level compared to your targets? Any treatment required with antibiotics? Early dry-off? Reproduction and breeding planning?
And what about automatically diverting low quality milk (high somatic cells or other problem) into a separate bulk milk tank?

Contact us for more info or for an offer

Know what’s going on, any time, any where.
We are performing field tests on various robot models, with the new Ekomilk Horizon Essential and AMP system (July-December 2017). Click here for an introduction to Ekomilk Horizon and AMP.
The system will be available from around Mars 2018 (exact timing depends on your region). The Ekomilk-AMP system operates fully autonomously and results, warnings and critical errors are sent over SMS, email and an online app (you can configure which notifications you get and how). Every analysis costs around $0.04 and accuracy figures compared to official lab results are (very) high. Click here for examples about realistic correlation coefficients for SCC, fat, protein and solid content.
How does it work?
First put your robot into “Sampling” (control, Shuttle) mode, by following the procedures prescribed by your robot manufacturer (first pause the robot etc). Instead of connecting the Shuttle or any other brand like Ori Collector or VMS sampler, Ekomilk will be connected. More accurately: Instead of the Shuttle (including its internal “spoon” or sampling head), the special electronic sample cup (like an electronic vial) of Ekomilk Horizon for milking robots, will be connected to the milk sampling line of the robot. This should work with all models of milking robots which work with the ORI Collector (ORI Sampler for milking robots) and which can create a certain milk sample volume without needing external components. (for instance in the case of GEA robots, an extra adaptor kit could be needed. Please refer to the connection of the ORI Collector to understand what is needed for your robot. From Q4 2017, there will also be a partnership program which might allow you to test Ekomilk-AMP on your robot under advantageous conditions. Please contact us for more info.

For Ekomilk-AMP, minimum 10ml is needed but between 30ml and 60ml is recommended for better accuracy. Please use the same milk sample tube as provided with your Shuttle (Ori Collector, VMS sampler) to connect your robot to the Ekomilk electronic sample cup. Please follow the same connection procedure and cleaning procedure for the sampling tube as prescribed by your Shuttle (Ori Collector, VMS sampler). The Ekomilk-AMP system does not need the air pressure. Please check with your robot supplier whether you can leave the air pressure outlet unused during sampling. If any doubts, please contact us or your robot supplier.

Photo of Ekomilk-AMP for robotic milking (AMS, VMS)
Please find here an indicative photo simulation of Ekomilk next to an old milking robot.

Ekomilk - AMP direct connection to milking robot (simulation)
Ekomilk – AMP direct connection to milking robot (simulation)

In yellow you can see the (here symbolic) milk tube which connects Ekomilk-AMP with the milking robot. This tube should be the same official type (ICAR approved) as used with the Shuttle or ORI sampler (ICAR approved). On the above image, indicated in white colour on the Ekomilk, there is a special digital sample cup. This special “robot” cup contains sensors to monitor the connection between milking robot and Ekomilk. Together with this cup comes also a detection and rinsing mechanism which allows automatic regular rinsing of the system , especially at the end of the day or after longer time intervals. The robot version of Ekomilk-AMP includes an extra package which includes all these essential components (without, the Ekomilk-AMP system will not be able to work automatically with the robot).

Maintenance and repair
There is a special version of the online remote error diagnostics AMP module and the Ekomilk AMP maintenance module included with each milking robot version of Ekomilk Horizon. These modules will monitor both your robot connection to Ekomilk-AMP and your Ekomilk instrument continuously. You can receive notifications via SMS, email and app when an error has to be solved or when maintenance is required.

Ekomilk&AMP availability
The robot (AMS, VMS) version of Ekomilk Horizon Essential (end-user mature prototype) can be shipped from March 2018. Orders can be placed from January 2018. Ekomilk Horizon Essential only measures SCC (somatic cell count).
From Q2 2018, the version of Ekomilk Horizon Unlimited for robotic milking (end-user mature prototype) will also be available and will measure SCC (>40 000 cells/ml), fat, protein and total solids content besides providing acidosis and ketosis indicators. The measurement accuracy of Ekomilk Horizon Unlimited is higher than Ekomilk Horizon Essential for all parameters (milk ingredients, components) measured because a multitude of parameters (+8 sensors) are combined online.

R&D partnership program
Ekomilk&AMP keep adding features and improving accuracy of milk analysis thanks for AI (self learning algorithms).
For robots, a partnership program will be launched to develop a milk sampler which works in “real-time”, during the milking, without needing the Shuttle (control, sampling) mode (which causes delays). We warmly invite all questions and suggestions.

Configure here your own Ekomilk-AMP solution
Click here if you don’t like long forms or if you want us to call you back

Observations of milking robots (voluntary milking systems, VMS) in practice
Milking robots are also called “voluntary milking systems (VMS)”. I noticed that in practice, such “voluntary” system also poses challenges, especially during the first weeks of introducing the new milking system on the farm. The herd has to adapt and not every individual (dairy cow) is as flexible. I have seen cows which cannot get enough from the new milk box and want (voluntary) to stay in it for quite some time – meanwhile blocking access for other cows who truly want to be milked (voluntary). Maybe that specific milking robot was too easily distributing that lovely easily palatable cow feed (the RFID transponder could maybe block cow feed distribution once the same cow comes in within a short interval). Also at the entry gate of the voluntary milking system, some cows get (voluntary) stuck, requiring intervention of the dairy farm staff. Probably there are small, transitional problems which also depend on cow characters.

An aspect of differentiation between voluntary milking system is the robotic arm (or robot arm). The robot arm provides the core of the automation of a milking robot. More observations soon.

Accuracy results from on-farm Ekomilk-AMP measurement for somatic cells, fat and protein content in cow milk

From March till June 2017, 2 evaluation users have used Ekomilk Horizon Unlimited on 4 dairy farms in Europe. For an introduction on the Ekomilk – AMP (Animal Monitoring) solution, please see this video (<3min):

In this post I give an overview of the accuracies and correlations achieved and the according on-farm test conditions.
As a brief summary, during official lab comparisons the average correlation coefficients were 98.7% for fat content, to 96% for SCC (somatic cell count), to 91% for protein content. In other words, over 90% accuracy can easily be achieved for each milk component or parameter measured once a full initial calibration is performed (also solids or dry-rest, conductivity, pH etc). Since the Ekomilk – AMP solution is improving every month, with more integration and automation options being added, it is realistic to say that a correlation coefficient of over 95% will be typical from 2018. Already the current 90-99% accuracy range seems more than sufficient for on-farm individual and bulk tank analysis.


On-farm testing procedures and conditions:

Of the 4 dairy farms visited, the farm size varied from 35 to 105 dairy cows. Cow breeds were Holstein, cross breds and a local breed which goes sometimes higher in fat content, over 6% (over 60 g/kg). Official ICAR approved samplers from the milking parlor manufacturer or from TruTest were used in order to create two representative and identical samples per cow (individual). The ambient temperatures varied from 17°C (early morning) to 28°C (sheltered room for milk analysis next to the milking parlor).
The individual milk samples were analysed by Ekomilk-AMP within 6 hours on average after the end of the milking process in the milking parlor. The official milk control labs are for instance the DHIA in the US and the Control Lechero in Spain and the Conseil Elevage (Controle Laitier) in France. Such official milk quality lab also analyzed each individual milk sample.

On the Cloud platform (AMP), a full calibration can be performed by the end-user. “Full calibration” means that the full measurement range which the official lab (i.e. a reliable, consistent reference method) can provide and which is physically possible for Ekomilk (for instance no butter or cream over 20% fat content unless diluted), can also be used for each milk component (parameter).

Milk components, measurement ranges and calibration:

The main milk components measured by Ekomilk-AMP and the according measurement ranges are (preliminary; custom ranges on request):

  • SCC (somatic cell count) from 40 000 cells/ml till 20 000 000 cells/ml,
  • F (fat content) from 0.1% till 12%,
  • P (protein content) from 0.5 till 9%,
  • S (solids, dry rest) from 4 till 24% (preliminary info).

“Full calibration” also means that the calibration is performed in the Cloud on multiple levels: per bulk tank, per farm and per animal if a consistent animal identification is present for consecutive (periodic) measurements. Combining these levels holds serious potential for further improvement of the accuracies.

While at this moment there is a manual intervention (doublecheck) implying a 2 day delay for a new calibration to be activated, by October 2017 this process will be completely automated thanks to concepts inspired by neural networks (NNets) and artificial intelligence (AI). The AMP calibration algorithms are self-learning and will in near-future better understand the influence of certain factors on the calibration – and hence improve the accuracy further and reduce the frequency of required calibrations. At this early stage, every 3 months, a full calibration (with more than 60% of the population i.e. almost all dairy cows on the farm sampled) is recommended in order to obtain similar accuracy results. Besides, at this early stage a quick partial calibration check with a few samples of the population (e.g. 10% of cows) is recommended every month for the most important milk components but this is only a temporary requirement. Studies are being conducted on the calibration frequency topic: the AMP self-learning algorithms (AI) will be able to reduce the recommended calibration frequency and the required number of samples (percent of the population).

Reliability of the measurements:

In 100% of the measurements, a wrong measurement on a milk component could be detected 100% reliably thanks to coherence checks which happen on tank, farm and animal level. In other words, there was never an error in the final measurement result which shifted the result into another measurement range. For these farms, 1082 physical measurements were available from analyzing composite (individual, 4 quarters) milk samples. The coherence checks excluded 7 measurements which were probably due to insufficient milk sample preparation (fat separation? see below) which was still manual (operator dependent) at this stage. In this context, it is important to mention the new AMPI Sample Feeder (SF). This automation accessory does not only automate; it also important because it verifies and prepares every milk sample. Depending on your SF model, the SF can:

  • de-aerate, removing air or gas included in the milk sample; this air removal is required for every milk sample which is to be measured within several hours after sampling.
  • level check: if the sample vial is empty or insufficiently filled, it will be detected automatically
  • mixing: butter fat or milk cream separation occurs already after 10 mins at room or lab temperatures around 24°C. The influence of fat separation
  • These farm trials were preceded by over 130 000 tests in the Ekomilk-AMP design labs and tests on other continents with earlier prototypes.


About using real-time results in the milking parlor and in the lab: the measurement results of Ekomilk-AMP, after applying the right calibration in the Cloud (AMP), can be received per SMS , via email, on the AMP web app on the browser or on any approved external database (DB) via the AMP APIs. In each AMP user account, it’s easy to add a few phone numbers or email addresses. Similarly, it’s just a click to share results with your preferred lab or vet so that your contact can follow in real-time, wherever he or she is located.

Laboratory comparisons, accuracy figures:

For these specific evaluations per farm, the Ekomilk-AMP results after calibration have been compared to the official lab results. The official labs often use analysis equipment from companies like Foss , Bentley or CEM. They have regular (more than once per 2 months) official calibration and certification procedures. In the graphs below, the correlation coefficients (more in general, coefficient of determination) are shown. Standard errors (like the theoretical standard deviation) are also summarized. RMSE figures are available in certain cases; RMSE were usually doublechekcs for our way of working using different algorithms from NNets or AI.
After lab comparisons, the average correlation coefficients varied from 98.5% for butter fat content, to 96% for SCC (somatic cell count), to 91% for milk protein content. In other words, over 90% accuracy can easily be achieved for each milk component or milk parameter measured. Less important ones such as solids or dry-rest, conductivity, on request also pH etc in any case achieve very high accuracy as well. Over 95% is realistic as a minimum for all milk components in mid-term (2018 if not earlier) for this new Ekomilk – AMP solution. Roughly summarising the parallel developments going on for Ekomilk – AMP, we could say  that every 3 months, the accuracy of Ekomilk – AMP is improving significantly. The examples below are only indicative; for a guarantee about the accuracy please contact us.

Selective dry cow therapy (SDCT) with AMPI and Ekomilk

For an introductory and abstract description of the Ekomilk – AMP system, click on the button:
Ekomilk-AMP system description

AMPI provides a special SDCT (Selective dry cow therapy) package which includes an accurate and fast on-farm Ekomilk somatic cell counter at $0.04 per test and a Cloud-based decision module for drying off and tele-vet module for vet advice on distance (remote diagnosis).

This SDCT package is the result of several years of research and field tests with major partners from the dairy industry (veterinaries, pharmaceuticals, udder hygiene, milk recording organisations or MROs, genomics and genetic selection organisations etc). This special SDCT package of course includes the new Ekomilk Horizon somatic cell counter at $0.04 per test with high accuracy and the optional auto-sampler.

Rear view of dirty dairy cow udder for selective dry cow therapy with AMPI and Ekomilk

Introduction to selective dry cow therapy (SDCT).

Selective dry cow therapy (SDCT) is the newer dry cow therapy following the traditional “blanket dry cow therapy” (BDCT). In the case of blanket dry cow therapy, every quarter (teat) of the udder is treated with antibiotics at drying off. As a reminder, “drying off” means that the daily milking routine of the lactation period is stopped; this is the start of the “dry cow period”. The dry cow period lasts till calving (parturition).

The idea behind blanket dry cow therapy (BDCT), practised for many years, was to cure clinical and even subclinical mastitis cases while also preventing new mastitis infections during the dry-off.

Drying off properly.

One of the challenges with the dry period is to manage the transition for the cow properly: if high yield milking is suddenly stopped, pressure in the udder builds up in the absence of regular milking. Instead the milking can be reduced gradually, less abruptly. As a reminder, in several countries, the minimum required daily milk yield for being profitable is around 25l per day, on average; of course we also have to take into account the typical lactation curve. Such high yield has to be reduced gradually for udder pressure control.

This control of udder pressure is relevant for mastitis control: usually mastitis infections occur when bacteria enter the teat canal. With udder pressure increasing, milk leaking can occur and this in turn means that the teat canal is not sealed against bacteria.

Adapting the cow’s nutrition is one of the ways to gradually reduce milk production.

Selective dry cow therapy (SDCT)

In the case of selective dry cow therapy (SDCT), a cow is only treated if certain udder health conditions are (not) fulfilled. Sometimes even udder quarters are distinguished in which case antibiotics treatments can even be more localized with a greater reduction of antibiotics use.

Selective dry cow therapy (SDCT) is now a legal obligation in several developed countries. For voluntary SDCT, the mindset of vets and farmers has to change, as several studies point out, for instance these by Prof Scherpenzeel:

One of the ideas of SDCT is to reduce the dependence of the dairy sector on antibiotics (antimicrobials).

Criteria for selective dry cow therapy (SDCT)

As Dr Supre from Flanders Milk Control Centre, Belgium, suggested, the following 3 criteria could be used for effective SDCT (compared to BDCT):

  • no clinical mastitis in that lactation,
  • daily milk yield had to be below 15kg and
  • the SCC for the three days prior to drying off had to be less than 100 000 cells/ml for heifers or below 150 000 cells/ml for cows.

Actually, as I heard from vets in other countries and as Dr Supre also mentioned, another study on SDCT showed that 300 000 cells/ml and 250 000 cells/ml are often adequate thresholds (cut-offs).

While earlier studies showed that SDCT is not as effective as BDCT, this is not true anymore in more recent studies. If properly managed with reliable tools and protocols, then SDCT can be at least as successful as BDCT. Administering antibiotics to low cell count cows at dry off would even entail other risks.

More details can be found on:


For an introductory and abstract description of the Ekomilk – AMP system, click on the button:
Ekomilk-AMP system description

AMPI special SDCT package

AMPI and Ekomilk special package for selective dry cow therapy SDCT

In the AMPI special SDCT package are included:

  • Ekomilk Horizon, the automated, accurate somatic cell counter for milk at $0.04/test with optional auto-sampler and sample feeder and optional measurement of fat, proteins, solids. Via AMPI a special discount price applies to Ekomilk instruments:
  • Cloud-based herd management software with tele-vet (remote diagnosis) module for deciding on how to dry-off each animal (usually dairy cows) with optional native Android app and with optional external database integration service with customised decision algorithms for your protected private usergroup.
    • Default decision algorithm, configurable by the user:
      • Clinical mastitis history of last 10 months
      • Choose your SCC (somatic cell count) threshold:
        • per udder quarter or composite
        • from 90 000 to 800 000 cells/ml
        • last milk yield
    • This default algorithm can be customised an extended with your preferred data, for instance:
      • bacteriology results
      • on-farm culturing (OFC) results
      • import from external databases
      • other
      • our IT team is at your disposal
  • Optional vet advice on distance over Skype, phone or email. Famous vet teams from several countries around the world are collaborating with us for instance the M-Team from Ugent in Belgium. Please contact us for details.


Available from Q4 2017. Please contact us for more details.


Automatic sample feeder sample changer autosampler liquids milk blood urine acqueous ampi H5

AMPI Auto-sampler: Automatic sample feeder and sampler robot for liquids

AMPI Auto-sampler: Automatic sample feeder or sample changer and sampler for liquids

The Animal Monitoring Platform is proud to announce the launch of the affordable automatic sample feeder and sampler robot for liquids such as milk, water, blood or urine or any other aqueous samples. The first model of the auto-sampler which will be launched is “AMPI H5” which has various, customisable features.

Introductory video of the Ekomilk AMPI Sample Feeder

What is it, what does it do?

Each AMPI automatic sample feeder is in its most basic version an automatic sample changer. Such automatic sample changer allows to test a tray or rack of  between 40 to 100 sample vials (sample cups, pots, tubes) sequentially. Such rack or tray of samples is a two dimensional plane where the sample vials (sample tubes, cups, pots) are positioned. The sample probe (also called sample needle or sample aspiration or sucking tube) of the automatic sample changer will move from sample to sample (i.e. sequentially) each time a signal (instruction) is received from an external source (or automatically, after a given time delay, optional).

How many samples (maximum)?

The exact amount of samples which can be tested in one tray or rack, depends on the size of the sample vials (also called sample cups or sample pots). The default is now fixed on 49 samples but customisation is possible. A rack or tray of 216x216mm (prelim. specs, can change without prior notification, please contact us) can be placed on the device. As a concrete example, these dimensions lead to around 10 by 10 sample vials (100 sample vials in total) in case the vials have an 18mm diameter and 2mm spacing on the rack  (tray) between the vials (cups, pots).

Of course we can also make larger size upon request. Please contact us via the online form so we understand your exact need.

Sample probe with or without pump?

The automatic sample feeder operates via an aluminium or the plastic sample probe (also called sample needle or sample aspiration / sucking tube) which moves from sample to sample. Optionally a peristaltic pump for sucking the liquid from each sample vial (or sample cup or pot) can be included. Such optional peristaltic pump also enables automatic rinsing of the tubes and probe besides enabling stirring and homogenising of the liquids before actual measurement starts (sample preparation).

Configure your own sample feeder
Click here if you don’t like long forms or if you want us to call you back

Sampler versus sample changer / sample feeder.

Finally, with peristaltic pump and in the optional reverse mode, the device can also operate as a true sampler by injecting an exact dose of liquid into each sample vial (sample cup).

Which signals or instructions are needed?

Via a signal (a string in your preferred format) on a communication port of the automatic sample feeder (USB, RS232, Wifi, Bluetooth, CANbus) , the sample probe (also called sample needle or sample sucking tube) of the sample feeder will move to the center of the next sample position in order to access the acqueous liquid (milk, blood, urine, other) in that sample vial.

The default instructions (strings) which are needed over RS232 will be specified in the coming months. Please contact us in the meantime; we usually customise according to your needs.

Who customises, who programs?

The sample vials are positioned on a two dimensional plane (basically a rack or tray of samples). In our factory, we program the sample vial positions depending on your sample cup (sample vial, pot) size. Common sample vial sizes are for instance 32mm as upper diameter. The next sample position will therefore be fixed at about 36mm from the previous sample position. The advantage of us performing the customisations is that we can fully test the system and we are used to customise.

Characteristics of “AMPI H5”:

In October 2017, we will launch the “AMPI H5” with the following preliminary characteristics:

  • End-user pricing starts from €2470 EXW per sample feeder unit for water, urine, milk or blood (aqueous liquids) testing and goes up to €4390 Mastitis: EXW (excluding upfront costs for customisation and for on-demand certification)
  • Low maximum operating speed, max. 3000 samples/day
  • Operation at room temperature
  • No certificates besides TUV or SGS for electronics, no FDA approvals (on request)
  • USB or RS232 communication for instructions from and to your instrument, sensor, equipment or computer
  • Dimensions approx. 450x450x300mm, around 15kg total
  • 110VAC or 240VAC, confirm during ordering
  • Sample vials (cups) positions with cups or vial size from 10 to 80mm can be pre-programmed in our factory with according rack or tray, for a minimum ordering volume (MOQ, nr of devices ordered).
  • Vertical movement of aspiration / sucking tube can be pre-programmed in our factory from 10 to 50mm
  • Integrated liquid (milk, blood, urine, water) level gauge to detect empty or insufficient samples
  • Automated recalibration menu option for two dimensional positioning system
  • 2 rinsing positions

These preliminary specifications can change at any time without prior notice. The prelim. drawing shown is only an illustration, not final. Final specifications will be confirmed on a best effort basis once you place your order.

Configure your own sample feeder
Click here if you don’t like long forms or if you want us to call you back

Optional features, future AMPI Sample Feeder models:

On future models, such as the high-end model “AMPI H5” we can include extra features. Please indicate on the online quote request form which optional features are important to you:

  • integrated peristaltic pump for sucking liquids, for injecting liquids (with sampler mode) and for auto-rinsing (with separate rinsing cups)
  • Wifi connection
  • Bluetooth connection
  • Can-bus
  • temperature sensor
  • liquid stirrer and homogeniser (mechanical)
  • flow and blockage detection
  • speaker for audio signal if crucial errors occur
  • advanced level sensor for gauging the level of liquid in each vial or cup
  • barcode scanner, RFID reader
  • integrated thermal microprinter
  • integrated LCD display
  • 12VDC power supply option (for operation in truck or car)
  • sampler mode: reverse mode for injecting an exact dose of liquid in each sample vial (sample cup)
  • feeder and sampler programming software for creating your own sample feeding and liquid sampling modes (+€690 EXW per 3y license  per user)
  • full sample preparation options: dilution, stirring, heating, cooling


A specific implementation of the AMPI H5 sample changer is used for Ekomilk Horizon.The AMPI H5 Ekomilk model has a specific set of features and functionalities, custom-made (tailored) for Ekomilk Horizon. Ekomilk Horizon allows somatic cell count and physical / chemical composition analysis in milk at about $0.04 per test.

Click here for more info about Ekomilk Horizon.

The link to the online  quote request (RFQ, offer request) form can be found here.