Sensor arrangement and method for operating a sensor arrangement

Abstract

A sensor array (100) with one or more sensors (201-204), with one or more operating means (101-104), each of which labels an object (801, 802), and with a processing unit (501) that is connected to the sensor or the sensors (201-204) via a communication connection (401), wherein each sensor (201-204) is designed for reading and unambiguously identifying an operating means (101-104). In each sensor (201-204), the time at which an operating means (101-104) was detected, the operating means identification of the operating means (101-104), a sensor identification that unambiguously identifies the sensor, and a quality variable q(t, s, b) determined based on time, sensor and operating means for the detection of the operating means (101-104) is sent to the processing unit (501). The invention further relates to a corresponding method.

Claims

1. A sensor array (100) with one or more sensors (201-204), with one or more operating means (101-104) each of which labels an object (801, 802), and with a processing unit (501), which is connected to the sensor or sensors (201-204) via a communication connection (401), wherein each sensor (201-204) is designed for reading and unambiguously identifying an operating means (101-104), characterized in that in each sensor, the time at which an operating means (101-104) was detected, the operating means identification of the operating means (101-104), a sensor identification unambiguously identifying the sensor (201-204) and a quality characteristic variable q(t, s, b) determined for detection of the operating means (101-104) dependent upon time, sensor and operating means, are sent to the processing unit (501) as sensor data, and that sensor quality values q(t, s) and operating means quality values q(t, b) are separated from measured quality characteristic variables q(t, s, b) by means of a matrix factorization algorithm implemented in the processing unit (501).

2. The sensor array (100) according to claim 1, characterized in that a fitness value h(T, s) is determined for a preset observation time interval T for each sensor (201-204), which specifies the probability of the sensor quality value q(t, s) determined at this sensor (201-204) being greater than a threshold value S.sub.s, and a fitness value h(T, b) is determined for each operating means (101-104), which specifies the probability of the operating means quality value q(t, b) determined for this operating means (101-104) being greater than a threshold value S.sub.b.

3. The sensor array (100) according to claim 2, characterized in that the sensors (201-204) and/or operating means (101-104) for which the fitness values h(T, s), h(T, b) are determined, and/or observation time intervals T and/or threshold values S.sub.s, S.sub.b, can be preset.

4. The sensor array (100) according to claim 2, characterized in that improvement potentials are determined from a comparison of fitness values h(T, s), h(T, b).

5. The sensor array (100) according to claim 2, characterized in that reproducibility values are determined from fitness values h(T, s), h(T, b).

6. The sensor array (100) according to claim 2, characterized in that action instructions are determined from fitness values h(T, s), h(T, b).

7. The sensor array (100) according to claim 2, characterized in that machine-readable control commands are generated from fitness values h(T, s), h(T, b).

8. The sensor array (100) according to claim 2, characterized in that parameterization instructions for sensors (201-204) are generated from fitness values h(T, s), h(T, b).

9. The sensor array (100) according to claim 8, characterized in that the parameterization instructions are only transmitted to those sensors which, with regard to their functionality, correspond to those sensors (201-204) from the fitness values of which the parameterization instructions are obtained.

10. The sensor array (100) according to claim 8, characterized in that parameterization instructions determined for sensors (201-204) of a plant are transmitted to sensors of a different plant.

11. The sensor array (100) according to claim 2, characterized in that remaining service lifetimes of sensors (201-204) and/or operating means (101-104) are determined from fitness values h(T, s), h(T, b).

12. The sensor array (100) according to claim 1, characterized in that it has multiple sensors (201-204) provided at different locations, and that a path analysis of an operating means (101-104) along the sensors (201-204) is performed in the processing unit (501).

13. The sensor array (100) according to claim 1, characterized in that additional measurement values can be separated from the quality characteristic variables, aside from the sensor quality values q(t, s) and operating means quality values q(t, b).

14. The sensor array (100) according to claim 1, characterized in that environmental sensors are provided for acquiring additional measurement values.

15. The sensor array (100) according to claim 13, characterized in that the additional measurement values are correlatable to the sensor quality values q(t, s) and the operating means quality values q(t, b).

16. The sensor array (100) according to claim 1, characterized in that the operating means (101-104) are designed as markings.

17. The sensor array (100) according to claim 16, characterized in that the operating means (101-104) are designed as codes.

18. The sensor array (100) according to claim 1, characterized in that the sensor or each sensor (201-204) is an optical sensor.

19. The sensor array (100) according to claim 18, characterized in that the optical sensor is a code reader.

20. The sensor array (100) according to claim 1, characterized in that the sensors (201-204) are arranged on a plant, wherein work processes are performed by means of the plant with the objects (801, 802).

21. The sensor array (100) according to claim 20, characterized in that the processing unit (501) is integrated into a plant control (701) of the plant or is integrated in a computing unit assigned to the plant.

22. The sensor array (100) according to claim 1, characterized in that the processing unit (501) is integrated in a cloud or in an OS-level virtualization system.

23. The sensor array (100) according to claim 1, characterized in that the processing unit (501) is integrated into one of the sensors (201-204).

24. The sensor array (100) according to claim 1, characterized in that the processing unit (501) has a data acquisition unit (503) for storing sensor data and a calculation unit (504) for calculating the quality values q(t, s), q(t, b).

25. The sensor array (100) according to claim 24, characterized in that derived characteristic values are calculated from the quality values q(t, s), q(t, b) in the processing unit (501).

26. The sensor array (100) according to claim 1, characterized in that an operating unit (601) is assigned to the processing unit, which operating unit (601) is designed for input of input values and output of output values.

27. The sensor array (100) according to claim 26, characterized in that sensor quality values q(t, s) and operating means quality values q(t, b) and/or characteristic values derived therefrom are displayed at the operating unit (601).

28. The sensor array (100) according to claim 27, characterized in that the development over time of quality values q(t, s), q(t, b) and/or characteristic values derived therefrom are visualized at the operating unit (601).

29. The sensor array (100) according to claim 26, characterized in that the operating unit (601) has an internet interface.

30. A method for operating a sensor array (100) with one or more sensors (201-204), with one or more operating means (101-104) each labeling an object (801, 802), and with a processing unit (501) that is connected to the sensor or the sensors (201-204) via a communication connection (401), wherein each sensor (201-204) is designed for reading and unambiguously identifying an operating means (101-104), characterized in that in each sensor (201-204), the time at which an operating means (101-104) was detected, the operating means identification of the operating means (101-104), a sensor identification unambiguously identifying the sensor (201-204) and a quality characteristic variable q(t, s, b) dependent upon time, sensor and operating means determined for the detection of the operating means (101-104) are transmitted to the processing unit (501) as sensor data, and in that sensor quality values q(t, s) and operating means quality values q(t, b) are separated from measured quality characteristic variables q(t, s, b) by means of a matrix factorization algorithm implemented in the processing unit (501).

31. The method according to claim 30, characterized in that a matrix M.sub.ij is formed for the matrix factorization algorithm, wherein one index i or j refers to the sensors (201-204) of the sensor array (100) and the other index j or i refers to the operating means (101-104) of the sensor array (100), that quality characteristic values q.sub.ij are determined for the matrix M.sub.ij during a calibration operation with the sensors (201-204) of the sensor array (100), that an ansatz for the dependency of the quality characteristic values on the variables q.sub.s, q.sub.b corresponding to the sensor quality values q(t, s) and the operating means quality values q(t, b) is selected in the form of a function ƒ (q.sub.s, q.sub.b), and that the function values of this function are fitted to the matrix values of the quality characteristic values q.sub.ij determined during the calibration operation.

32. The method according to claim 31, characterized in that the function values are fitted by means of an optimization method.

33. The method according to claim 32, characterized in that the optimization method contains an error minimization.

34. The method according to claim 31, characterized in that ƒ (q.sub.s, q.sub.b)=q.sub.s.Math.q.sub.b is selected as the ansatz.

35. The method according to claim 31, characterized in that AI (artificial intelligence) algorithms are used to calculate the function values of the function.

36. The method according to claim 30, characterized in that a path analysis is performed for each operating means (101-104) such that a chronological sequence is produced of which sensors (201-204) the operating means (101-104) were detected by, wherein based on the chronological sequence, a path is produced, which path is composed of individual path segments, which path segments are composed of pairs of sensors (201-204), which respectively have detected the operating means (101-104) successively.

37. The method according to claim 36, characterized in that the absolute frequency at which this path segment was traveled during the time interval is determined for every path segment of the path of an operating means (101-104).

38. The method according to claim 37, characterized in that the paths and/or the frequencies of traveled path segments are visualized.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention is explained below based on the drawings. They show:

(2) FIG. 1: A schematic depiction of an exemplary embodiment of the sensor array according to the invention.

(3) FIG. 2: A block circuit diagram of the sensor array according to FIG. 1.

(4) FIG. 3: A first variant of a further development of the sensor array according to FIG. 2.

(5) FIG. 4: A second variant of a further development of the sensor array according to FIG. 2.

(6) FIG. 5: A diagram with fitness value for the sensor array according to FIGS. 1 and 2.

(7) FIG. 6: A diagram with improvement potential for the sensor array according to FIGS. 1 and 2.

(8) FIG. 7: An example of a path analysis for the sensor array according to FIGS. 1 and 2.

(9) FIG. 8: A depiction of a range of angular positions for a sensor of the sensor array according to FIGS. 1 and 2.

(10) FIG. 9A: A depiction of angular positions for sensors of the sensor array according to FIGS. 1 and 2.

(11) FIG. 9B: A correlation of the angular positions according to FIG. 9A with quality value.

(12) FIG. 10: An example of an OEE (overall equipment effectiveness) calculation for a sensor of the sensor array according to FIGS. 1 and 2.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(13) FIG. 1 shows a schematic depiction of an exemplary embodiment of the sensor array 100 according to the invention. FIG. 2 shows a block circuit diagram of this sensor array 100.

(14) The sensor array 100 comprises an arrangement of sensors 201-204, which in the present case are designed as optical sensors in the form of code readers. The code readers can be designed in the form of scanners or camera sensors. The sensors 201-204 are arranged at preset workstations of a production line 301. Objects 801, 802 with which work processes are performed or which are required to perform work processes, are conveyed along the production line 301. Each object 801, 802 is unambiguously labeled with an operating means 101-104 in the form of a code. When an object 801, 802 is located in the reading range of a sensor 201-204, the code applied to the object 801, 802 can be read.

(15) Of course, the topology of the sensor array 100 as well as the number of sensors 201-204 and operating means 101-104 are not limited to those shown in the depictions in FIGS. 1 and 2.

(16) As shown in FIG. 2, the sensors 201-204 are connected to a processing unit 501 via a communication connection 401. The processing unit 501 comprises a memory unit 502, a data acquisition unit 503 and a calculation unit 504. It is advantageous for the memory unit 502 to be integrated in the data acquisition unit 503.

(17) The processing unit 501 is connected to an operating unit 601, which is designed for data input and output.

(18) FIGS. 3 and 4 show two variants of an extension of the sensor array 100 from FIG. 2 such that a plant control 701 of a plant, which plant control 701 controls the production line 301, is integrated. In the arrangement from FIG. 3, the communication connection 401 is used for connecting the plant control 701. In the arrangement from FIG. 4, a control connection 402 is provided as a second communication channel for connecting the plant control 701. This second communication channel uses a hardware or software interface of its own, for example, such that the second communication channel uses the same connection as the first communication channel, such as Ethernet, wherein different software connections (ports) are used, however.

(19) According to the invention, upon detection of an operating means 101-104 by means of a sensor 201-204, both the operating means identification as well as the own sensor identification and a quality characteristic variable q(t, s, b) determined upon detection of the operating means 101-104 are sent as sensor data to the processing unit 501 and stored in the memory unit 502.

(20) The quality characteristic variable q(t, s, b) is a time-dependent value and is a function of parameters of the sensor 201-204 as well as parameters of the read operating means 101-104.

(21) This quality characteristic variable q(t, s, b) comprises properties of the operating means 101-104, i.e., codes, especially their reflectivity, the contrast of the operating means 101-104, the number of read operations of the sensors 201-204 required for detection of the operating means 101-104, and, if applicable, necessary error corrections upon evaluation of read operations or values derived therefrom.

(22) According to the invention, a sensor quality value q(t, s) and an operating means quality value q(t, b) are obtained in the processing unit from the quality characteristic variable by means of a matrix factorization algorithm, i.e., separate quality values are obtained for each sensor 201-204 and each operating means 101-104. According to the invention, it is thereby achieved that the qualities of the sensors 201-204, on the one hand, and the qualities of the sensors 201-204, on the other hand, can be acquired individually and independently from one another.

(23) The sensor quality values q(t, s) and the operating means quality values q(t, b) can be displayed to a user by means of the operating unit 601, wherein especially also the development over time of these quality values can be displayed.

(24) Moreover, a fitness value h(T, s) is determined for each sensor 201-204 for a preset observation time interval T, wherein the fitness value specifies the probability that the sensor quality value q(t, s) determined at this sensor 201-204 is greater than a threshold value S. In addition, a fitness value h(T, b) is determined for each operating means 101-104, wherein the fitness value specifies the probability that the operating means quality value q(t, b) determined for this operating means 101-104 is greater than a threshold value S.sub.b.

(25) A user can preset the observation time interval and the threshold values S.sub.s, S.sub.b via the operating unit 601.

(26) The fitness values provide a measure for the operational fitness of the sensors 201-204 and operating means 101-104.

(27) FIG. 5 shows examples of such fitness values for the sensors 201-204. The fitness values are scaled to values from 0% to 100%, wherein the value 100% corresponds to full, unrestricted operational fitness and the value 0% corresponds to completely insufficient operational fitness.

(28) As shown in FIG. 5, the sensor 201 has the lowest fitness value and the sensor 204 has the highest fitness value.

(29) These fitness values can be visualized at the operating unit 601.

(30) Improvement potentials can be derived from the fitness values. To do so, the highest fitness value is taken as a reference and the difference relative to this reference is determined for the remaining fitness values.

(31) FIG. 6 shows the result of the improvement potentials derived from the fitness values from FIG. 5. The sensor 204 has the highest fitness value, which serves as reference. With this, the improvement potentials depicted in FIG. 6 for the sensors 201-204 are obtained, which improvement potentials can also be visualized by means of the operating unit 601.

(32) As is evident from FIG. 6, the sensor 201 has a high improvement potential. This alerts the user to a need for action, for example, such that the sensor 201 must be replaced. The other sensors 202, 203 have low improvement potentials, such that they can be further operated without additional measures.

(33) To generate a measure for the reproducibility of the fitness values and/or improvement potentials, the spread for the fitness values can be calculated.

(34) Action instructions as well, especially including control commands, can be generated based on the fitness values, which action instructions can be transmitted in machine-readable form to the plant control 701 and directly executed there.

(35) Moreover, parameterization instructions can be derived from the fitness values for sensors 201-204 for these or similar sensors.

(36) Finally, remaining service lifetimes for sensors 201-204 can be determined based on fitness values.

(37) A path analysis can be produced for the sensor array 100 that specifies which operating means 101-104 are detected by which sensors 201-204 at which times. The results of such path analyses can be visualized by means of the operating unit 601.

(38) FIG. 7 shows the result of such a path analysis for a sensor array 100 with the sensors 201-204, for which the fitness values according to FIG. 5 were also calculated.

(39) FIG. 7 shows the path of a product labeled with an operating means 101 through the plant. Here, the path of the operating means 101 is plotted over a preset time interval (in this case, from 3:00 μm to 3:12 pm).

(40) As is evident from FIG. 7, the operating means 101 switches multiple times between sensors 201 and 204 over a short time, which indicates a problem for the product in the work process. Moreover, it is evident that the sensors 202, 203 are not used at all, such that no load is present for them during the observation period.

(41) In general, the matrix factorization algorithm can be extended such that additional measurement values can be prepared from the quality characteristic variables q(t, s, b) determined by direct measurement, such as influences due to ambient light.

(42) Alternatively or additionally, environmental sensors, such as temperature sensors, can be provided for acquiring additional measurement values.

(43) These measurement values can be correlated with the sensor quality values and operating means quality values.

(44) An example of such a measurement value is the angular position of an operating means 101 within a scanning beam E of a sensor 201, as is depicted schematically in FIG. 8. The angular position can vary within the angular range oc.

(45) FIG. 9a shows the variation of the angular positions. These angular positions can be contrasted with the corresponding sensor quality values or the fitness values to visualize irregularities. The result is shown in FIG. 9b.

(46) As is shown as an example in FIG. 10, the sensor quality values q(t, s) and operating means quality values q(t, b) determined for the sensor array 100, as well as characteristic values derived therefrom, can be drawn on for complex evaluations, such as KPI (key performance indicator) evaluations, by means of which OEE (overall equipment effectiveness) characteristic values can be calculated, which represent measurement figures for the overall plant effectiveness of plants. These evaluations can also be visualized at the operating unit 601.

(47) In FIG. 10, F represents the number of successful scans, i.e., successful reads, during an observation period A (operating time) of an operating means 101-104 by a sensor 201-204. D and E are the total number of scans performed with the sensor 201-204 during the observation period. C is a user input and defines a target quantity per unit time, extrapolated to the observation period. The run time B corresponds to the difference of the observation period minus the time during which there are no measurements by the sensor. The quality factor F/E corresponds to the fitness value h(t, s) of the sensor 201-204 that was calculated for a threshold value S.sub.s, where especially S.sub.s=0.

(48) From this are derived quality losses as efficiency losses and output losses as effectiveness losses due to minor stops or reduced speeds for conveying products as well as availability losses due to plant standstill times.

LIST OF REFERENCE NUMERALS

(49) (100) sensor array (101-104) operating means (201-204) sensors (301) production line (401) communication connection (402) control connection (501) processing unit (502) memory unit (503) data acquisition unit (504) calculation unit (601) operating unit (701) plant control (801, 802) objects E scanning beam T observation time interval