EP4189443A1

EP4189443A1 - Cloud platform for underground water reservoirs

Info

Publication number: EP4189443A1
Application number: EP21763405.4A
Authority: EP
Inventors: Michail MAVROFORAKIS
Original assignee: Inttrust SA Information Technology Trust
Current assignee: Inttrust SA Information Technology Trust
Priority date: 2020-07-30
Filing date: 2021-07-28
Publication date: 2023-06-07
Also published as: WO2022024009A1; GR20200100453A

Abstract

A Cloud platform including i) memory means, ii) communication means configured to receive data from a plurality of IoT devices, each IoT device associated with a distinct well, and output means configured to provide output to each IoT device, which output is related to the underground water reservoir associated with said IoT device, and iii) processing unit configured to analyze data employing cognitive techniques, such as Artificial Intelligence and Machine Learning techniques and optionally statistical analysis techniques and/or signal analysis techniques and configured to provide trends of the behavior of water parameters of the underground water reservoirs, for example under pre-defined usage scenarios. The Cloud platform and a plurality of IoT devices establish a system for monitoring the properties of underground waters.

Description

Cloud platform for underground water reservoirs

[0001] The invention refers to a computer-implemented method to monitor and analyze underground waters locally and in a plurality of underground water reservoirs. The invention further refers to an apparatus and Information Technology (IT) tools for the implementation of the method.

[0002] Existing methods monitor on an ad-hoc and sparse basis the underground waters level depth of a well at a specific geographic location, as well as some relevant aspects of the quality of a well’s water or the degree of its contamination. In general, existing methods analyze an underground water reservoir independently from other reservoirs, rather than revealing information regarding dependencies on the qualities and behavior of a plurality of reservoirs.

[0003] The object of the invention is to provide a computer-implemented method for the systematic, continuous, on-line, real-time monitoring of the properties of underground waters. A further object of the invention is a computer-implemented method for modeling the behavior of underground waters under environmental and/or users’ actions in local, regional and global geographic areas. A further object of the invention is a computer- implemented method for the prediction of the behavior of underground waters under usage and environmental scenarios. Further, the invention aims in a Cloud-based ecosystem that is used for improving the sustainability of underground water systems. The invention aims also in an apparatus for the implementation of the computer-implemented inventions. [0004] The invention is defined in the independent claims.

[0005] Dependent claims define further features that add further advantages to the invention.

[0006] The computer-implemented method of the invention is used for the on-line, real-time monitoring and analysis of the properties of a plurality of underground water reservoirs, each one of the plurality of underground water reservoirs associated with a distinct well. The method comprises:

• installing an loT device including sensors to each of said plurality of reservoirs and/or to the associated well, for monitoring data related to usage parameters, water parameters and environmental parameters, at various time instants

• registering the loT devices to a Cloud platform equipped with a processing unit applying cognitive techniques, such as Artificial Intelligence and Machine Learning techniques, input means configured to receive data from said plurality of loT devices, output means configured to provide output related to said plurality of underground water reservoirs and a memory unit

• transmitting data monitored by registered loT devices to the Cloud platform

• storing transmitted data in the memory unit of the Cloud platform

• processing the data transmitted to the Cloud platform by said processing unit, so as to infer inter-dependencies/correlations between the usage parameters, the water parameters and the environmental parameters of the plurality of underground water reservoirs.

[0007] The term Cloud platform used in the present application encompasses generic client-server platforms in www or local. A “plurality” means more than one.

[0008] Water parameters describe the quantity/capacity and quality of the water in the underground water reservoirs. Examples of such parameters are the Static Water Level (SWL), i.e. the distance from the ground of the surface of underground waters local reservoir, water depth, the quality, temperature, pH, electrical-conductivity, Total Dissolved Solids (TDS), Oxidation- Reduction Potential (ORP), Dissolved Oxygen (DO). [0009] The water depth measurements in rest, i.e. when the installation does not pump, may be tracked and analyzed in long term, typically in weekly or monthly aggregations throughout a year or more, in order to estimate statistics and infer trends that describe the status and capacity of the underground water reservoir in the greater region.

[0010] The water depth measurements during instability, i.e. during pumping or during refill, may be tracked and analyzed in short term, typically at the resolution of few minutes or finer, in order to estimate the volatility, dynamics and adaptive statistics that describe the dynamics and drainage- recovery profile of the underground water reservoir, as well as the performance of the pumping infrastructure. [0011] Usage parameters are pumping parameters and controlling parameters. Pumping parameters relate to the parameters that describe the pumping of the water from the underground water reservoir. Examples of such parameters are Pumping Water Level (PWL), pumping times and durations, Drawdown (DD), time of water level stabilization, pumping and refill rate, pumped quantity. Controlling parameters relate to loT device health-check, diagnostics and process-optimization parameters.

[0012] Environmental parameters relate to the environmental conditions. Examples of environmental parameters are atmospheric pressure, temperature, humidity, wind parameters, rain gauge data, solar power and UV radiation, dew point, soil moisture, soil pH.

[0013] Examples of Cognitive techniques, e.g. Machine Learning techniques that may be applied by the processing unit of the Cloud platform, are neural networks, support vector machines (SVMs), decision trees, random forests etc. The processing unit of the Cloud platform may further apply statistical analysis techniques and/or signal analysis techniques. [0014] The method may use data analytics via Cognitive techniques, e.g.

Artificial Intelligence and Machine Learning techniques to provide predictions about, but not limited to: Static Water Level (SWL), Pumping Water Level (PWL), anomaly detection (water reservoir extreme conditions), fault detection (hardware / sensor failures), predictive maintenance (prevent failures, assisted calibration). [0015] Parameters like the SWL, PWL, water quality parameters, loT device’s working values (e.g., efficiency rate of the air compressor, electricity consumption), along their history, are used to determine the current status of the reservoirs and predict the future status, for example under specific usage scenarios, including pumping scenarios and cultivation patterns.

[0016] The Cloud platform may process the data transmitted by the loT devices to define parameters for controlling the loT devices, process optimization and/or pumping parameters of the underground water reservoirs to be employed in the usage of the underground water reservoirs and transmitting said parameters to the loT devices.

[0017] Optionally at least an loT device that is used in a method according to the invention, includes a local memory unit and the method includes storing and processing the data monitored by the loT device to the local memory unit of the respective loT device -“Edge computing”. The data transmitted by the loT device may be raw data or data that have been monitored and further processed - including data generated from the raw data that are monitored - by the loT device.

[0018] In an embodiment of a method of the invention, at least an loT device includes a local processing unit and the method includes processing locally the data that is monitored by the loT device by the local processing unit of the respective loT device to derive at least a usage parameter and/or an environmental parameter and/or a water parameter reflecting the status and the behavior of the respective underground water reservoir. [0019] In a further embodiment of the method, at least an loT device includes surveillance means of the well and/or the loT device.

[0020] Based on the tracked statistics during pumping and no-pumping cycles, the performance and accuracy of the loT device can be re-assessed and re-calibrated periodically, for example every month, when the loT device installation, e.g., maintenance, or environmental factors, e.g. summer/winter, change. Moreover, if the quantity and/or quality of the water tend to drop, relevant warnings are issued and changes to the pumping schedule are proposed.

[0021] The data monitored and transmitted by the loT devices may include a time series of a parameter. [0022] According to the invention an loT device for monitoring the properties of an underground water reservoir comprises sensors to monitor time series of a parameter of an underground water reservoir and means to transmit and receive data to and from a Cloud platform. [0023] The loT device of the invention may include a memory unit and a processing unit configured to store and process respectively the time series.

[0024] The invention suggests a Cloud platform including i) memory means, ii) communication means configured to receive data from a plurality of loT devices, each loT device being associated with an underground water reservoir, and having output means configured to provide output to each loT device, which output is related to the underground water reservoir associated with said loT device, and iii) processing unit configured to analyze data employing cognitive techniques, such as Artificial Intelligence and Machine Learning techniques to provide trends of the behavior of water parameters of the underground water reservoirs, for example under pre-defined usage scenarios.

[0025] The Cloud platform may include a neural network or other Artificial Intelligence and Machine Learning techniques having an input layer configured to receive a dataset including data from said plurality of loT devices of the plurality of reservoirs.

[0026] The Cloud platform and a plurality of loT devices establish a system for monitoring the properties of underground waters.

[0027] According to the invention a computer-implemented method, which provides usage parameters, in particular pumping parameters, for a plurality of distinct wells, applies a cognitive technique and comprises: training the cognitive technique by data-sets, whereby each dataset contains data obtained from a plurality of loT devices, each loT device being associated with a distinct well of said plurality of wells.

[0028] Optionally the computer-implemented method includes a neural network with an input layer configured to receive said-data sets. [0029] The computer-implemented method may be used for mapping a water table of a predefined geographical area.

[0030] A device according to the invention operates a plurality of distinct pumping installations and includes a learning machine, e.g. a neural network, which provides pumping parameters and/or process optimization parameters to said installations. The learning machine comprises an input stage configured to receive datasets, whereby each dataset contains data obtained from a plurality of loT devices, each loT device being associated with a distinct well.

[0031] The processing means of the Cloud platform includes particular means that may offer at least some of the following:

• providing optimal pumping and controlling/function parameters for process and operation optimization, dynamic control of the loT device

• providing quantity, quality and sustainability trends of a plurality underground water reservoirs

• predicting the depth of an underground water table, in order to make suggestions for drill location of new wells

• providing diagnostics and predictive maintenance support for the loT device and the pumping equipment

• supporting farming techniques and decision related to optimized water usage, management and waste reduction, at local and/or larger, e.g., municipal or governmental or global scale

• supporting medium- and long-term agricultural planning that depends on the underground waters availability and quality • providing accurate data for large scale map representation of underground waters, thus improving information accuracy, availability and transparency

• providing an ecosystem for the development of applications (from third parties) related to underground waters

• identifying subtle correlations and impact of environmental parameters, e.g., rain and snow height, of the broader geographic areas to underground waters reservoir level

• computing the impact of the pumping quantities and rate of neighboring wells registered to the Cloud platform to the water level and quality of a specific well.

[0032] Each well may be associated with a pumping device or it may be used purely for other purposes, for example research and monitoring.

[0033] With a method or system according to the invention it is possible to monitor and analyze in real time or near real time, on a periodic basis, underground waters both locally and in a plurality of underground water reservoirs. The method or the system can be employed for providing a large- scale simulation of the behavior of underground water systems representation and understanding their trends, as well as improving their sustainability.

[0034] Description of embodiments of the invention follow: [0035] An example of a system according to the invention includes a plurality of loT devices. Each loT device includes sensors, hardware with a memory unit, a processing unit and input means and output means. The sensors are appropriate to monitor

• water parameters, such as the Static Water Level (SWL), i.e. , the distance from the ground of the surface of underground waters local reservoir, the quality, temperature, pH, electrical-conductivity, Total Dissolved Solids (TDS), Oxidation-Reduction Potential (ORP), Dissolved Oxygen (DO)

• pumping parameters, such as Pumping Water Level (PWL), pumping times and durations, Drawdown (DD), time of water level stabilization, pumping rate, pumped quantity

• Controlling parameters, such as loT device health-check parameters, diagnostics and process-optimization parameters

• local environmental parameters, such as atmospheric pressure, temperature, humidity, wind parameters, rain gauge data, solar power and UV radiation, dew point, soil moisture, soil pH

[0036] Parameters are measured directly by the sensors or deduced from measured data. The data, including the directly measured parameters, are measured over time with a steady or variable frequency. The measured data may form data series, for example water level vs. time, which subsequently are used in the analysis.

[0037] Each loT device is installed to a distinct reservoir. Distinct underground waters reservoirs are reservoirs from which water is pumped via distinct water pump systems, such as wells. In general, these reservoirs may have solid boundaries therebetween. It may be that distinct reservoirs are indirectly connected, for example through subterranean rivers. [0038] The monitoring parameters, such as the frequency of the measurements may be adapted so as to optimize the energy consumption of the loT device, as well as the amount and value of data stored and processed. The parameters are selected using data analysis and ML. [0039] The measurements are stored and processed locally in the memory unit of the loT device and, consequently, uploaded to the Cloud. Appropriate means, software and/or hardware means, process the data to obtain computed parameters that reflect the condition of the respective reservoir. These means include Cognitive techniques, such as ML and Al techniques as well as other in-house developed and state of the art algorithms and/or semi-conductors.

[0040] Parameters that are measured directly or induced from the measurements may include parameters related to the reservoir’s water capacity, the limits and rate of drainage and refill of the well along their seasonality, the induced information regarding the consistency, porosity and permeability of the well’s lower rocky surroundings.

[0041] The loT devices are registered to and interact with a Cloud-based Software as a Service platform (SaaS). The platform is equipped with modules that process the measurements and the derived parameters, which are inputted from the loT devices.

[0042] Measurements that are obtained from the loT devices and respective derived parameters are forwarded to the Cloud platform either real-time or at pre-defined or ad-hoc, for example in case of communication problems’ situations, intervals. The data received from the loT devices are processed by the Cloud platform to provide insights and visual representations of the information. They may be also used to train the Al and ML modules in order to provide trends related to the parameters and qualities of the underground waters, such as the sustainability and seasonality of the well attributes and qualities and of the corresponding reservoir. Once trained the Al and/or ML modules may be employed to suggest optimal configuration and usage values, e.g. pumping times and duration, optimal irrigation parameters, i.e. times, duration, quantities, preferable cultivation conditions, such as weather, soil conditions in case of water usage for agricultural cultivations, to identify if two or more wells pump from the same or connected reservoirs - for example by computing the probability that the water of two or more nearby wells belong to the same underground waters reservoir. Further interpolation, extrapolation and ML techniques are also used to provide clues for the sustainability of the quantity and quality of the underground water reservoirs, according to expected or assume (what-if) scenarios, based on environmental and usage parameters and, thus, support the relevant decision making. The process of the data by ML and Al techniques detect and reveal dependencies between the data of the wells of a reservoir of the plurality of the reservoirs and the data of the others of the plurality of the reservoirs. The Al and ML modules of the Cloud platform are trained by records of datasets obtained from the loT devices. The records include data from a plurality of loT devices, so as to reveal hidden dependencies between the plurality of the reservoirs.

[0043] Once trained, the ML and Al modules of the Cloud platform may be employed to simulate the behavior of the reservoirs under specific usage and environmental scenarios, so as to determine the appropriate parameters for future usage. The appropriate usage parameters are transmitted via appropriate means to the loT devices and/or to the users of the system. The Al and ML modules of the Cloud platform may define the effect that a usage scenario of a single reservoir or of more than one reservoir may have on the plurality of reservoirs that are registered to the platform. [0044] The Cloud platform may also include a module to process the data received to provide several visualization aspects, including a geographic map depicting the underground water reservoir upper levels and other relevant attributes and qualities, e.g., water quality parameters and trends.

Claims

1. Computer-implemented method for on-line, real-time monitoring and analysis of the properties of a plurality of underground water reservoirs, each one of the plurality of underground water reservoir associated with a distinct well, comprising:

• transmitting data monitored by registered loT devices to the Cloud platform

• storing transmitted data in the memory unit of the Cloud platform · processing the data transmitted to the Cloud platform by said processing unit, so as to provide dependencies between the usage parameters, the water parameters and the environmental parameters of the plurality of underground water reservoirs.

2. Computer-implemented method according to claim 1, whereby the

Cloud platform processes data transmitted by the loT devices to define parameters for controlling the loT devices and/or usage parameters of the underground water reservoirs to be employed in the usage of the underground water reservoirs and transmitting the said parameters to the loT devices.

3. Computer-implemented method according to any one of claims 1 or 2, whereby at least an loT device includes a local memory unit, the method including storing the data monitored by the loT device to the local memory unit of the respective loT device.

4. Computer-implemented method according to any one of the preceding claims, whereby at least an loT device includes a local processing unit, the method including processing locally data that is monitored by the loT unit by the local processing unit of the respective loT device to derive parameters reflecting the status and the behavior of the respective underground water reservoir.

5. Computer-implemented method according to any one of the preceding claims, whereby at least an loT device includes surveillance means of the loT device and/or the well.

6. Computer-implemented method according to any one of the preceding claims, whereby the data monitored and transmitted by the loT devices include a time series.

7. Computer-implemented method according to any one of the preceding claims, including a memory unit and a processing unit configured to store and process respectively time series.

8. Computer-implemented method according to any one of the preceding claims, whereby the processing unit of the Cloud platform further applies statistical analysis techniques and/or signal analysis techniques.

9. A Cloud platform including i) memory means, ii) communication means configured to receive data from a plurality of loT devices, each loT device being associated with a distinct underground water reservoir, and having output means configured to provide output to each loT device, which output is related to the underground water reservoir associated with said loT device, and iii) processing unit configured to analyze data employing Cognitive techniques, such as Artificial Intelligence (Al) and Machine

Learning (ML) techniques, to provide trends of the behavior of water parameters of the underground water reservoirs.

10. A Cloud platform according to claim 9, whereby the processing unit is configured to provide trends of the behavior of water parameters of the underground water reservoirs under pre-defined usage scenarios.

11. A Cloud platform according to claim 9 or claim 10, whereby the cognitive technique includes an input stage configured to receive a dataset including data from said plurality of loT devices of the plurality of reservoirs.

12. A Cloud platform according to any one of claims 9 to 11, whereby processing unit is configured to analyze data employing statistical analysis techniques and/or signal analysis techniques.

13. A system for monitoring the properties of underground waters comprising a Cloud platform according to any one of claims 9 to 12 and one or more loT devices, each loT device associated with a distinct well, comprising sensors to monitor time series of a parameter of an underground water reservoir and means to transmit and receive data to and from a Cloud platform.

14. A computer-implemented method, which applies a cognitive technique and comprises: training the cognitive technique by data-sets, whereby each data-set contains data obtained from a plurality of loT devices, each loT device being associated with a distinct well, and providing usage parameters, in particular pumping parameters, to each loT device of said plurality of loT devices.

15. A computer-implemented method according to claim 14, including a neural network with an input layer configured to receive said-data sets.

16. A computer-implemented method according to claim 14 or claim 15 including mapping a water table of a predefined geographical area.

17. A device to operate a plurality of distinct wells, including means to execute the computer-implemented method according to any one of claims 14 to 16.