CLEANING PROCEDURES IN A FILLING MACHINE"

❖ ❖ ❖

TECHNICAL FIELD

The present invention relates to an automated monitoring apparatus and method for monitoring cleaning procedures in a filling machine.

STATE OF THE ART

Filling machines are known, which are typically used to fill containers, such as bottles or small bottles made of plastic or glass, cans, jars or the like, with a pourable product; generally, these filling machines are installed in a production plant, which comprises further machines, such as labelling machines, capping machines, packaging machines etc.

Filling machines basically comprise a carousel rotatable around a vertical axis, a tank containing the pourable product, and a plurality of filling units or valves, which are carried peripherally by the carousel, are connected to the tank by means of respective filling circuits or ducts, and are conveyed by the carousel along a transfer path substantially shaped as an arc of a circle.

Generally, the above-mentioned machines comprise an inlet device, typically an inlet star-wheel configured to sequentially supply empty containers to the carousel, and an outlet device, typically an outlet star-wheel configured to receive full containers from the carousel and transfer them to another part of the plant (for example, to a labelling machine).

The carousel typically comprises a plurality of support elements each adapted to receive and maintain in a vertical position, below each filling valve, a respective container to be filled.

Each filling valve is adapted to supply a pre- established volume of pourable product to the corresponding underlying container, while it is conveyed along the transfer path, thanks to the rotary movement imparted to it by the carousel.

In some cases, the pourable product supplied to the containers is a food product, for example, natural or carbonated water, tea, or the like; in other cases, the pourable product supplied to the containers is not a food product, for example, soaps, detergents, room fragrances, or the like.

It is a known requirement for these filling machines to clean the filling circuits and ducts and the filling valves periodically and/or during certain operating events, such as, for example, a product change (for example, from water to a different beverage or to a different product), in order to ensure hygienic conditions and the absence of contamination due to residual product from previous processing.

In particular, the filling machines may be cleaned by using a cleaning product, such as a detergent solution (containing chemical agents, for example soda, acid, disinfectant, in addition to water), which is made to circulate in the machine instead of the pourable product intended to fill the containers, in order to remove any residual product and/or any biological or bacteriological contaminants.

In order to limit consumption of detergent and improve the cleaning, the use of cleaning systems of the so-called Cleaning in Place (CIP) type is known, namely which are arranged locally with respect to the same filling machine.

These cleaning systems generally comprise at least one cleaning tank, storing the cleaning product, and to which appropriate pumping means are coupled; and a fluidic circuit that connects this cleaning tank with the tank of the pourable product and with the filling valves. This fluidic circuit furthermore defines a return path from the filling valves to the same cleaning tank in order to create a closed recirculation path, for recovery of the cleaning product.

In a known manner, CIP cleaning procedures may envisage successive cleaning cycles including various cleaning steps, for example washing, rinsing, waiting, etc, and also various types of cleaning products, for example, water, soda, etc. Each cleaning step may also envisage specific operating conditions, for example in terms of a corresponding duration, execution temperature, types and product concentrations, etc.

The cleaning procedures are therefore rather complex and various factors may influence their efficiency and correct result.

The present Applicant has also realized that the above- mentioned cleaning procedures may influence parts of the filling machine, such as ducts, seals, etc., and their wear (for example, due to the use of caustic substances); however, today there are no effective solutions that offer operators the possibility to monitor the cleaning procedures and evaluate and quantify their effects on the filling machine.

WO2016025248A1 discloses a method for monitoring a CIP procedure previously performed. The method includes accessing CIP procedure data stored in a CIP library. The CIP procedure data includes CIP phase data associated with a plurality of CIP phases in the CIP procedure previously performed, equipment data associated with objects used in the CIP procedure during one or more of the CIP phases, and consumable data associated with consumables consumed by corresponding equipment in the CIP procedure.

OBJECT AND SUMMARY OF THE INVENTION The aim of the present invention is to provide a solution for monitoring the cleaning system of the filling machine, allowing to solve the above-mentioned problem.

According to the invention, this aim is achieved by a monitoring apparatus and method, as defined in the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the present invention, some preferred non-limiting embodiments thereof will be described in the following, purely by way of example and with the aid of the attached drawings, wherein:

- Figure 1 is a general block diagram of a filling machine, configured to fill containers with a pourable product, and of an associated CIP-type cleaning system;

- Figure 2 shows a reference model for the time trend of a cleaning parameter during a cleaning procedure carried out by means of the cleaning system;

- Figure 3 graphically shows identification of possible anomalies during a cleaning procedure, by means of comparison between a real-time trend of a cleaning parameter and the reference model; and

- Figure 4 is a block diagram of a procedure for determining reference cleaning models associated with the cleaning procedure.

DETAILED DESCRIPTION With reference to Figure 1, a filling machine is generally designated with 1; the filling machine is configured to fill, with a pourable product, for example a pourable food product, a plurality of containers, such as bottles, jars, small bottles made of plastic or glass, or cans.

In a known manner, here not described in detail, the filling machine 1 comprises:

- a product supply tank 2, configured to store the pourable product intended to fill the containers; and

- a plurality of filling valves 4; these valves are, for example, coupled to a carousel, which is rotatable so as to feed the containers along a filling path, and are configured to fill the containers while they are fed along the filling path.

In particular, each filling valve 4 is hydraulically coupled to the product supply tank 2 by means of a supply circuit 6, so as to selectively supply a predetermined amount of pourable product to at least one container at a time, while the same container is conveyed along the filling path thanks to the rotary motion imparted to it by the carousel.

The filling machine 1 further comprises, installed locally at the same filling machine 1, a cleaning system 10, of the Cleaning in Place (CIP) type. The cleaning system 10 is configured to implement a cleaning procedure of the filling machine 1, and in particular of at least the product supply tank 2, the supply circuit 6, and the filling valves 4.

In detail, the cleaning system 10 comprises: a washing tank 12; and a cleaning circuit 14 that hydraulically connects the washing tank 12 with the product supply tank 2 and therefore with the filling valves 4 through the above- mentioned supply circuit 6. The washing tank 12 stores a cleaning product. The cleaning product may be, for example, a detergent or washing fluid. The cleaning tank 12 may be coupled to pumping means.

The cleaning system 10 further comprises a return circuit 16, that defines a hydraulic return path from the filling valves 4 towards the same washing tank 12, so as to create a closed recirculation path for recovery of the cleaning product.

The filling machine 1 comprises a controller 18, that is coupled to the filling valves 4 to control and supervise general operation thereof and to receive information on a corresponding operating state; in a manner not illustrated, the same controller 18 may furthermore be operatively coupled to the cleaning system 10.

According to an aspect of the present invention, an automatic monitoring apparatus 20 is operatively coupled to the cleaning system 10. The automatic monitoring apparatus 20 is configured to perform monitoring of the cleaning procedures. The cleaning system 10 is configured so that each cleaning procedure is carried out by means of a flow of the cleaning product. The cleaning system 10 is configured so that each cleaning procedure is carried out by means of recirculation of this flow through the supply tank 2, the supply circuit 6, the filling valves 4, the return circuit 16, the washing tank 12 and the cleaning circuit 14.

A set of cleaning parameters comprises at least one of the following variables: flow temperature of the cleaning product, a type of cleaning product, a flow speed of the cleaning product, a concentration (for example, a concentration in volume) of a chemical component in the flow of the cleaning product.

This set of cleaning parameters could, for example, comprise only the above-mentioned temperature.

The monitoring apparatus 20 is configured so that monitoring of each cleaning procedure comprises, for each cleaning parameter of this set, detection of a time trend of the parameter.

The monitoring apparatus 20 is configured so that monitoring of each cleaning procedure comprises, for each parameter, determining any anomalies, by comparing the detected time trend with a respective reference model.

The subsequent Figures 2-4 refer to an exemplary case in which the detected cleaning parameter is the above- mentioned temperature.

The monitoring apparatus may monitor the cleaning procedures by detecting information that may be useful to the user in order to improve regulation of the cleaning system 10, so that subsequent cleaning procedures may be improved or more easily and/or intuitively optimised.

The reference model is defined by a plurality of cleaning steps.

In Figure 2 the model is designated by TM. A first step s1, a second step s2 and a third step s3 are, for example, shown in the same Figure 2. Each cleaning step is associated with a respective range of values of the cleaning parameter and with a respective time window. In Figure 2: the first step si is associated with a first window wl and with a first range r1; the second step s2 is associated with a second window w2 and with a second range r2; the third step s3 is associated with a third window w3 and with a third range r3.

The monitoring apparatus 20 is configured so that detection of an anomaly corresponds, for example, to: a value of the parameter that is outside the respective range, for at least one instant that is within the respective time window; or a value of the parameter that is within the respective range, for at least one instant that is outside the respective time window. With reference to Figure 3, the detected trend of the parameter is shown with a dotted line, and is designated by TT. The first anomaly al is determined by a value of the parameter that is outside the first range r1, for one instant that is within the first window w1. The third anomaly a3 is determined by a value of the parameter that is outside the third range r3, for one instant that is within the third window w3. The second anomaly a2 is determined by a value of the parameter that is within the first range r1, for one instant that is outside the first window w1. The fourth anomaly a4 is determined by a value of the parameter that is within the third range r3, for one instant that is outside the third window w3.

The monitoring apparatus 20 is configured to calculate, for each cleaning procedure, a respective numerical index (I _CIP )• This numerical index is indicative of the effectiveness and/or quality of the cleaning procedure. The monitoring apparatus 20 is configured to calculate this numerical index based on an integrative evaluation over time of the mechanical, physical and chemical actions associated with the cleaning procedure. In particular, the mechanical action is represented, in this evaluation, by the flow speed of the cleaning product. In particular, the physical action is represented, in this evaluation, by the temperature of cleaning product flow. In particular, the chemical action is represented, in this evaluation, by the concentration of at least one chemical component in the cleaning product flow.

The monitoring apparatus 20 is configured to calculate the numerical index by numerically estimating the following formula: where t indicates any time instant during the cleaning procedure, v is the flow speed of the cleaning product, cone is the concentration of at least one chemical component in the flow of the cleaning product, and Temp is the temperature of the cleaning product flow (for example, cone is a concentration in volume).

The monitoring apparatus 20 is configured to determine, for each cleaning procedure and based on the final value of the numerical index calculated, at least one corrective action of a cleaning recipe. The cleaning recipe may be defined by the time trend of the speed and/or by the time trend of the temperature and/or by the concentration. This corrective action may comprise a change of the time trend of speed and/or a change of the time trend of temperature and/or a change of concentration.

The monitoring apparatus 20 is configured to allow a user to visualise, for each cleaning procedure, the final value of the numerical index; in this manner the user may, if necessary, autonomously determine this corrective action. In general, the corrective action may be aimed at improving the environmental sustainability of the cleaning procedure and/or improving the effectiveness of the cleaning procedure and/or reducing the impact of the cleaning procedure in terms of wear of the wearable parts of the cleaning system 10 and/or of the filling machine 1.

The monitoring apparatus 20 is configured to allow a user to visualise in real-time, during each cleaning procedure, the instantaneous value of the numerical index. In this manner, the user may undertake at least one corrective action before the end of the cleaning procedure, for example, by interrupting the cleaning procedure in order to use a smaller amount of cleaning product.

The monitoring apparatus 20 is configured to generate, for each anomaly detected, a warning representative of the anomaly detected.

As shown in the above-mentioned Figure 1, the monitoring apparatus 20 furthermore comprises: a processing unit 22; a sensor assembly 26; a memory 24; and a user interface 28. The processing unit 22 is in communication with the sensor assembly 26, with the user interface 28 and with the memory 24. The processing unit 22 may be part of the controller 18.

The processing unit 22 may comprise any suitable computing unit, for example a microprocessor, a microcontroller or a similar digital processing module that carries out a set of software instructions stored in a nonvolatile memory.

In a possible embodiment, the processing unit 22 may be implemented as an in-cloud network processing platform, accessible remotely with any suitable hardware and software equipment.

The memory 24 stores information associated with monitoring of the cleaning procedures and/or the reference models.

This memory 24 may be implemented by any non-volatile data storage unit; for example, the memory 24 may comprise a data base and may be implemented in an in-cloud processing platform .

The monitoring apparatus 20, for each cleaning procedure and for each parameter of the set of parameters, is configured to detect the time trend of the parameter by means of the sensor assembly 26 and the processing unit 22.

The monitoring apparatus 20, for each cleaning procedure and for each parameter detected, is configured to detect any anomalies by means of the processing unit 22 and possibly by means of the memory 24. The memory 24 is used for purposes of comparison with the reference model.

The sensor assembly 26 may comprise a plurality of sensors. This plurality of sensors may comprise a first sensor 26a. This plurality of sensors may comprise a second sensor 26b. At least one of the sensors is a temperature sensor, which is configured to detect a temperature of cleaning product flow. At least one of the sensors is a sensor of a type that is configured to detect information indicative of the type of cleaning product. At least one of the sensors is a speed sensor, which is configured to detect a flow speed of the cleaning product. At least one of the sensors is a concentration sensor, which is configured to detect information correlated with the concentration of at least one chemical component in the cleaning product flow.

The monitoring apparatus 20 is configured to display the final value and/or the instantaneous value of the calculated numerical index, through the user interface 28.

The monitoring apparatus 20 is configured to generate the warning relative to the anomaly through the user interface 28.

The monitoring apparatus 20 is configured to monitor the operation of the cleaning system 10 and recognise each time a cleaning procedure is implemented in the filling machine 1.

In this regard, the processing unit 22 may receive from the controller 18 data indicative of the start of a new cleaning procedure; alternatively or additionally, the same processing unit 22 may receive detection signals from the sensor assembly 26 and process the same detection signals in order to determine the start of the new cleaning procedure.

The processing unit 22 is therefore configured to analyse each cleaning procedure in order to verify the effectiveness and performance thereof, and furthermore to evaluate any anomalies or deviations with respect to the desired or reference model.

The processing unit 22 may therefore warn the operator of this anomaly in real-time, by means of an appropriate alarm or signalling generated through the user interface 28; alternatively or additionally, the processing unit 22 may cause autonomously and in an automated manner the implementation of appropriate corrective actions for the cleaning procedure, for example by changing a duration of a cleaning step or at least a value of the corresponding cleaning parameter.

For each cleaning procedure, the processing unit 22 may furthermore check if for some reason the procedure is interrupted before it is fully executed, so as to be able to recognise the anomaly and consequently change a counter of executed cleaning cycles (preventing it from being incremented).

According to an aspect of the present invention, the above-mentioned numerical index (I _CIP ) is used by the processing unit 22 to contribute to detection of the above- mentioned anomalies, and in particular to recognise whether or not the cleaning procedure is impacting on the working life of the wearable parts of the filling machine 1, such as for example, seals, membranes, etc.

Consequently, the processing unit 22 may be configured to generate in real-time, namely during execution of the cleaning procedure, an alarm or warning, through the user interface 28 or another communication system (for example, e-mail, SMS), so as to notify of the anomalous situation observed in the cleaning procedure (for example, an increase of time to reach a high temperature required by a cleaning step).

Furthermore, the same numerical index (I _CIP ) may be used by the processing unit 22 to automatically determine new settings of the cleaning procedure, in essence by changing the cleaning "recipe", in order to optimise the cleaning process (for example a reduction or an increase of the duration of one or more cleaning procedures, a change of the concentration of the chemical products used, a change of temperature, etc). The recipe may be changed for example by varying the temperature and/or the speed of the cleaning product and/or the concentration of a chemical component in the cleaning product.

The above-mentioned changes may be aimed, for example, at improving the numerical index I _CIP and/or reducing consumptions, in particular of cleaning product. The changes automatically introduced in the cleaning procedure may also be shared by the processing unit 22 with the operator by means of the above-mentioned user interface 28, or by means of further and different communication tools (for example e-mail or SMS).

By way of example, in the case where the numerical index I _CIP calculated is lower than a theoretical reference quality index envisaged for a given cleaning procedure, the processing unit 22 may be configured to determine an extended duration or an increased temperature of one or more of the cleaning steps.

All the information (received from the filling machine 1 or calculated by the processing unit 22) are furthermore stored in the memory 24 so as to be available and usable for further a-posteriori analysis, such as for example: checking a cumulative duration (for example, in terms of the total number) of the cleaning procedures executed; checking a cumulative duration (for example, in terms of the number) of high-temperature or low-temperature cleaning steps; detailed analysis of each cleaning procedure performed, such as in terms of duration, number of cleaning steps, temperature, concentration of chemical products used, etc.; and estimation of the evolution of the number of cleaning procedures in the near future, with the possibility for the operator to plan in advance maintenance activities based on historical trends highlighted, and, for example, on planning suggestions proposed by the same processing unit 22.

In particular, the operator, based on the analysis of the numerical index I _CIP , may act to reduce the concentration of the chemical agents in the cleaning product, in order to achieve a greater sustainability of the cleaning procedures. Furthermore, the same operator may estimate the evolution of the number of cleaning procedures in the near future, in order to plan any maintenance operations.

The apparatus is configured to allow a user to visualise the trend of the numerical index in real-time, thus allowing the user to be able to act on the cleaning procedure in realtime. In this manner the cleaning procedure may be improved and/or made more sustainable in real-time.

It should also be highlighted that the reference models associated with the cleaning procedures may be stored in the memory 24, so as to be recalled when the corresponding cleaning procedure is identified, for example, at a pourable product change (a different reference model stored in the memory 24 being associated, in the example, with every product change).

According to a further aspect of the present solution, the monitoring apparatus 20 may furthermore be configured, for each cleaning parameter, to autonomously and automatically determine at least one reference model or the reference models.

For this purpose, the monitoring apparatus 20 is configured, for each parameter, to implement monitoring of a first plurality of cleaning procedures, in order to extract at least one reference model. The monitoring apparatus 20 is configured to extract this at least one reference model by comparing the time trends detected with one another for each of said first subsequent cleaning procedures and clustering the time trends together based on a mutual similarity. Each extracted model is based on a respective cluster of time trends.

In particular, the processing unit 22 may be configured to carry out a clustering of the time trends detected, based on similarity between the same time trends, and determine the reference models based on respective clusters of time trends.

Afterwards, the processing unit 22 may monitor a second plurality of cleaning procedures and determine any anomalies by means of comparison of the detected time trend for at least one cleaning parameter with the time trend of the most similar reference model.

In this regard, Figure 4 schematically and by way of example shows the step, designated by Ml, of monitoring the first plurality of cleaning procedures, with comparison of the detected trends of the at least one cleaning parameter (designated by TT1, TT2, TT3 and TT4); and the step, designated by EM, of extracting the models, by comparing the detected trends with one another, in the example with the identification of the models TM1 and TM2.

The same Figure 4 shows the step, designated by M2, of monitoring the second plurality of cleaning procedures, with detection of the detected trends of the monitored parameter, designated by TT5 and TT6; the step, designated by CM, of comparing the detected trends with the most similar model, in the example TM2 and TM1; and the consequent step of detecting any anomalies, designated by AD.

From an examination of the characteristics of the discussed solution, the advantages that it allows to obtain are evident.

In any case, it is again underlined that this solution allows to monitor the cleaning procedures in real-time or a- posteriori, by identifying any anomalies or deviations with respect to reference models. In particular, when anomalies are detected, warnings can be generated or appropriate corrective actions can be automatically implemented, in order to optimise the cleaning operations of the filling machine.

In general, the solution described allows, among other advantages: optimisation of maintenance, estimating the evolution of the cleaning procedures based on historical trends; detection of the deviations, by comparing the actual characteristics with the expected or historical ones, with the possibility of sending warnings to the operators in realtime, in order to restore the optimal conditions and ensure cleaning quality; optimisation of the cleaning procedures, in terms of time and costs; sustainability, allowing, for example, reduction of the chemical products used; saving, thanks to the reduction of the labour due to the automatic monitoring of the quality of the cleaning procedures and the reduction of downtimes, thanks to optimisation of the duration of the cleaning procedures.

Lastly it is clear that modifications and variations may be made to what has been described and illustrated here without thereby departing from the scope defined by the claims.

In particular, it is again underlined that the monitoring of the cleaning procedures may be based on any parameter relating to the same cleaning procedure, different from the above-mentioned temperature, to which reference has been made only as an example.