Computer-implemented process for processing harvest plans and associated hardware and systems

Abstract

One embodiment of a computer-implemented process is for processing a harvesting plan for harvesting natural resources from a target environment. The computing system performs a simulation process to process environmental data, operational data, and plan data to simulate conditions in the target environment after a predetermined first time period during which the harvesting plan is simulated to be implemented. The computing system also performs constraint processing to process the constraint data and the simulated conditions of the target environment to determine whether any of the environmental constraints are simulated to be violated. If none of the environmental constraints are simulated to be violated, the computing system flags the harvesting plan as compliant. The computing system also performs the simulation process and constraint processing on each of the respective plan data to generate multiple alternative harvesting plans and respective plan data and flag multiple compliant harvesting plans.

Description

[Technical Field]

[0001] The present invention relates to processes and associated apparatus and systems for use in the extractive industries. Embodiments of the invention find application in areas such as, but not exclusively, logging, fishing, land-based mining, and deep-sea mining. [Background technology]

[0002] Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is solely for the purpose of providing a context for the present invention and is not to be construed as an admission that any or all of such matters form part of the prior art or were common general knowledge in the art relevant to the invention as they existed in Australia or elsewhere prior to the priority date of this application.

[0003] Managing extractive operations typically requires a delicate balance between inherently conflicting considerations, such as maximizing productivity while keeping environmental impacts within limits established by regulatory agencies with jurisdiction over the target environment. Attempting such a balance will generally be greatly aided by a detailed understanding of the target environment, particularly the ability to predict and quantify the environmental impacts that may result from various aspects of the proposed extractive activities. However, in many target environments, such information may be rudimentary, scarce, or nonexistent. This presents particular challenges for entities wishing to initiate extractive operations in little-known target environments, such as, by way of non-limiting example, the deep ocean. Summary of the Invention [Problem to be solved by the invention]

[0004] It is an object of the present invention to overcome or substantially ameliorate one or more of the disadvantages of the prior art, or to provide a useful alternative.

[0005] In one aspect of the invention, a computer-implemented process is provided for processing a harvesting plan for harvesting natural resources from a target environment, the process including configuring a computing system to: access environmental data indicative of the target environment; access constraint data indicative of a plurality of environmental constraints applicable to the target environment; access operational data indicative of equipment deployed within the target environment; and access plan data indicative of the harvesting plan; after a predetermined first time period during which the harvesting plan is simulated to be implemented, perform a simulation process to process the environmental data, operational data, and plan data to simulate states of the target environment; perform constraint processing to process the constraint data and the simulated states of the target environment to determine whether any of the environmental constraints are simulated to be violated; flag the harvesting plan as compliant if none of the environmental constraints are simulated to be violated; and perform the simulation process and constraint processing on each of the respective plan data to generate a plurality of alternative harvesting plans and respective plan data and flag a plurality of compliant harvesting plans.

[0006] Preferably, the process includes defining metrics for quantifying the desirability of harvesting plans, calculating individual metrics for each compliant harvesting plan, and presenting the harvesting plan with the highest metric to a user of the computing system for approval after a predetermined second period of time has elapsed, the predetermined second period of time being shorter than the predetermined first period of time.

[0007] In one embodiment, if an environmental constraint is simulated as being violated with respect to the harvest plan, the computing system is configured to prompt a user of the computing system to select at least one of a plurality of predetermined potential revisions to the harvest plan.

[0008] Preferably, the simulation process uses a digital twin configured to simulate current operational conditions based on operational data input from sensors, and the digital twin is configurable to simulate future operational data that would be simulated if the harvesting plan were implemented.

[0009] Preferably, the simulation process uses probabilistic analysis to process operational data and future operational data to model causal relationships for multiple indicators of environmental impact.

[0010] In one embodiment, generating a plurality of alternative harvest plans and respective plan data includes incrementally modifying manipulated variables of the harvest plans to generate modified manipulated variables and processing the modified manipulated variables with an optimization algorithm.

[0011] Generating the multiple alternative harvesting plans and respective plan data may include randomly generating harvesting plan initiation parameters and processing the randomly generated initiation parameters through an optimization algorithm. Examples of such initiation parameters include a proposed harvesting start location within the target environment, a proposed harvest rate of the natural resource, a proposed efficiency level, and a proposed power usage level.

[0012] An embodiment of the computing system maintains a mode indicator that indicates an active adaptive management state or a passive adaptive management state. In this embodiment, when the mode indicator is in the passive adaptive management state, the generation of multiple alternative harvest plans is constrained within a safe region of the operating envelope, whereby the alternative harvest plans are likely to comply with environmental constraints. When the mode indicator is in the active adaptive management state, the generation of multiple alternative harvest plans is biased toward boundary regions of the operating envelope where the alternative harvest plans are close to or non-compliant with environmental constraints.

[0013] In one embodiment, the computer system switches the mode indicator between an active adaptive management state and a passive adaptive management state in response to user input. In another embodiment, the computer system is configured to calculate an uncertainty score associated with the probabilistic analysis, and when the uncertainty score falls below a threshold, the computer system recommends to the user to set the mode indicator to the active adaptive management state, and when the uncertainty score rises above the threshold, the computer system automatically sets the mode indicator to the passive adaptive management state.

[0014] In one embodiment, the harvesting plan is operatively implemented during a predetermined first time period, and multiple alternative harvesting plans and respective plan data are generated during the predetermined first time period. In one such embodiment, starting operating parameters of at least some of the alternative harvesting plans correspond to predicted operating parameters of the implemented harvesting plan at the elapse of a predetermined second time period.

[0015] Some embodiments of the process include defining a third time period, the third time period being longer than the second time period, and after the third time period has elapsed, the computing system is configured to match the empirically derived environmental data with the empirically derived operational data to assist research scientists and / or data analysts in applying probabilistic and / or machine learning techniques to the empirically derived environmental data and the empirically derived operational data to update at least one of the probabilistic analysis, constraint data, tolerance levels for state changes, and / or ecosystem models. In some such embodiments, the second time period may be between five days and six months, and the first and third time periods are each between one month and one year.

[0016] In some embodiments, the computing system is configured to maintain a portal accessible to regulators and / or the public, the portal making available at least one type of information: environmental data, constraint data, and operational data.

[0017] According to a third aspect of the present invention, there is provided a computing system configured to carry out the process as set out above.

[0018] According to another aspect of the present invention, there is provided a system for harvesting natural resources from a target environment, the system including a computing system configured to perform the process as described above, a harvesting device deployed in the target environment, and a plurality of sensors deployed on and around the harvesting device and in the target environment, the plurality of sensors communicating with the computing system via a telemetry link.

[0019] The features and advantages of the present invention will become more apparent from the following detailed description of preferred embodiments, given by way of example only, in conjunction with the accompanying drawings, in which: [Brief explanation of the drawings]

[0020] [Figure 1] 1 is a schematic diagram illustrating the architecture of one embodiment of a computing system according to the present invention; [Figure 2] FIG. 2 is a schematic diagram illustrating elements of environmental logic implemented in one embodiment of a computing system according to the present invention. [Figure 3] 1 is a flow chart illustrating passive adaptive management implemented in one embodiment of the present invention. [Figure 4] 1 is a flow chart illustrating active adaptive management implemented in one embodiment of the present invention. [Figure 5] FIG. 2 is a schematic diagram illustrating the architecture of a digital twin used in the embodiment of FIG. 1. [Figure 6] 1 is a flow chart illustrating active adaptive management implemented in one embodiment of the present invention. [Figure 7] 1 is a flowchart illustrating the process of generating a collection plan. [Figure 8]10 is a flowchart illustrating a process for using empirically derived data to update aspects of the system logic. [Figure 9] FIG. 1 is a side view of a schematic representation of a system for harvesting nodules from a target environment on the seabed. DETAILED DESCRIPTION OF THE INVENTION

[0021] Computing system 1 can be configured to perform the processes illustrated in Figures 3, 4, 6, 7, and 8. From a high-level perspective, the system architecture includes several communicatively interlinked modules, for example as shown in Figure 1.

[0022] Computing system 1 can be utilized in connection with a variety of different types of extractive industries, such as logging, fishing, land-based mining, and deep-sea mining. One preferred implementation of the present invention focuses on the deep-sea mining example shown schematically in Figure 9. However, those skilled in the art will understand that the general concepts of the present invention can readily be adapted for use in other types of extractive industries.

[0023] In an example of deep-sea mining, a system for harvesting nodules from the ocean floor at depths of approximately 4,000–4,500 meters requires multiple collectors 4 that roam the seafloor using jets of water to collect the nodules. Typically, this nodule harvesting process involves lifting the surface layer of seafloor sediments, likely causing a sediment plume to move within ocean currents. This provides a non-limiting example of the types of operational activities and associated environmental impacts that computing system 1 is configured to analyze and predict. Once collected by collectors 4, the nodules are transported by compressed air bubbles up a substantially vertical pipe called a riser 5, which delivers the nodules to a vessel 6 for subsequent processing.

[0024] In a typical embodiment, the modules of computing system 1 may be implemented in a cloud computing context, except that, as outlined in more detail below, sensors 3 are located in and around the target environment and workstations 21 are provided onboard vessel 6 to run the software necessary for the telemetry link between sensors 3 and the various modules of computing system 1 in the cloud. Some examples of publicly or commercially available software modules that may be used in a preferred embodiment are: · Kognitwin Digital Twin by Kongsberg, Computational Fluid Dynamics, provided by DHI Group Open source software for Bayesian networks, and / or · Database services provided by InfluxData or InfluxDB by Dremio, including:

[0025] Before the deep-sea sampling operation begins, the computing system must be given access to various types of data. In some embodiments, some or all of this data is stored in a locally accessible database 24. In other embodiments, some or all of this data is stored in a cloud computing data storage facility or is accessible to computing system 1 via subscriptions to third-party data provision services. This data includes environmental data that helps populate environmental logic module 8 and ecosystem model 9. When the process flow shown in FIG. 4 first proceeds to step 4.2, environmental data is collected. This typically includes environmental baselines established by collecting samples from the target environment. The samples are brought to a laboratory and analyzed to identify the composition of the ecosystem, including key species. The formulation of ecosystem model 9 allows for modeling of the food web of the target environment. Food web analysis involves consideration of nutrient inputs and outputs and interrelationships between various species, including predation, detritus, etc. This process generates environmental data indicative of the target environment to which computing system 1 has access. This provides a baseline against which changes in environmental conditions caused by human impacts from extractive activities can be measured.

[0026] In some contexts, aspects of the environmental data may be accessible to computing system 1 from the public domain or from commercial sources, such as subscriptions to third-party databases. In a deep-sea context, information about ocean currents may be made available in this manner.

[0027] The pre-production step also typically involves consultation with experts to formulate and quantify key ecosystem indicators, such as counts of various key species. In the context of deep-sea harvesting, examples of key ecosystem indicators are: Benthic plume deposition, · Definition of conservation areas; Plume dispersion, - number of major animal species, Noise generation, the number of major microbial species, ·eDNA analysis, Cultural and / or scientific DGM programs, ·Water chemistry, Heavy metal content of fauna, Organic flux, Nitrification concentrations in water and pore water, ·Pore water chemistry, Phytoplankton density / biomass, Nutrient concentration, ·Water chemistry, Plankton community composition, Sediment properties, Caste density, and Precipitation radiochemistry, which may include:

[0028] The individual components of environmental logic module 8 are shown in FIG. 2. In addition to ecosystem model 9, environmental logic module 8 provides computing system 1 with access to constraint data 10 that indicates a number of environmental constraints applicable to the target environment. This constraint data 10 is essentially a realistic representation of the environmental regulations applicable to the target environment and to which extraction activities are subject. It may include data indicating environmental aspects that must be tracked. Environmental logic also includes data indicating acceptable levels of environmental state change 12. In the context of deep sea mining, one example of the type of constraint that may be coded into the constraint data includes constraints regarding suspended and deposited sediments caused by the operation of collector 4, as shown below. [Table 1]

[0029] Another example of a type of constraint that may be applicable in a deep-sea mining context and that may be encoded in the constraint data is noise generated by the collection operation that propagates through the environment and may disrupt the ecosystem. Such a constraint may be that the noise level generated by operating the collector 4 and riser 5 does not exceed 75 dB, as measured by a far-field buoy spaced at a fixed distance from the operation site.

[0030] In the context of land-based mining, some examples of constraint data that may be coded into a computing system are: [Table 2]

[0031] As shown in FIG. 5, the digital twin 2 maintains a virtual representation of assets 17 configured to track the current operational state of physical assets used in the collection system, such as collectors 4, risers 5, nodule-receiving vessels 6, remotely operated vehicles 7, moorings, buoys, environmental monitors, etc. This is based on operational data input from sensors 3 deployed on and around the collection equipment and within the target environment. Examples of operational data include data such as equipment location, equipment speed, collection rate, efficiency, power consumption, material handling variables, logistics variables, etc. Some of the sensors 3 deployed within the deep-sea target environment are shown in FIG. 9.

[0032] The vessel 6 that receives the nodules is A return water sampler to sense the water quality, turbidity, oxygen level, redox level, pH, nitrate level, and nutrient level of the return water; CO, NO x ,SO2 and VOC sensing air pollution sensor, Sound level meters for measuring atmospheric noise levels, Electromagnetic flow sensors for measuring return water velocity, return water flow rate, return water CO2 level, and return water temperature; Multi-belt scales for measuring conveyor belt throughput, Global navigation satellite systems for monitoring heading, speed, x-location and y-location; and -Hydrophone for measuring underwater noise levels.

[0033] Riser 5, located in the center column position, Ultra-short baseline transponders for measuring x-position, y-position, and depth, Hydrophones for measuring underwater noise levels, and Includes CTD probes for measuring conductivity, temperature, and depth. The base of Ryza 5 is an underwater acoustic positioning transponder for measuring x-position, y-position, and depth; and -Hydrophone for measuring underwater noise levels.

[0034] Each collection machine 4 Underwater particle size analyzers for measuring turbidity and total suspended solids, a light sensor for measuring light levels; Hydrophones for measuring underwater noise levels, A power meter to measure power consumption, Densitometer for measuring discharge density, A flow sensor for measuring the discharge flow rate, Densitometer for measuring jumper hose density, Pressure sensors for measuring water pressure, Flow sensor for measuring jumper hose flow, Doppler velocity logger to measure the speed of the collector above the seabed, an inertial navigation system for determining the orientation of the collection vehicle; and -Has an underwater acoustic positioning transponder for measuring x-position, y-position, and depth.

[0035] Each remotely operated vehicle 7 is Bottled water sampler for measuring water quality, Turbidity sensors for measuring turbidity, An inertial navigation system for measuring heading, roll, and pitch; Underwater acoustic positioning transponders for measuring x-position, y-position and depth, CTD probes for measuring conductivity, temperature, and depth, Hydrophones for measuring underwater noise levels, and ·Has a light sensor for measuring light levels.

[0036] Each mooring device: Ultrasonic current meters for measuring ocean currents, speeds, and directions; Hydrophones for measuring underwater noise levels, Water samplers for measuring water quality, turbidity, oxygen levels, redox levels, pH, nitrate levels, and nutrient levels; and Includes a CTD probe for measuring conductivity, temperature, and depth. Each buoy is -Ultrasonic current meters for measuring ocean currents, speed, and direction.

[0037] The sampling system also uses satellite imagery to measure animal movement patterns and operating fleet discharge.

[0038] The sensors 3 communicate with the computing system 1 via a telemetry stream 22, whereby each of the sensors 3 communicates with a central control station 21 located on a vessel 4 floating near the equipment performing the sampling operation.

[0039] The digital twin also utilizes a scenario modeler 18 to simulate future operational data that is predicted to be appropriate for the harvesting equipment as the harvesting plan is simulated to be implemented.

[0040] The environmental logic module 8 also includes an impact analyzer 11 configured to use probabilistic analysis to process operational data and future operational data to model cause-and-effect relationships for multiple indicators of environmental impact, such as key ecosystem indicators. A preferred embodiment uses a Bayesian network for probabilistic analysis, configured to take the current state of the environment, ecosystem model, and proposed harvest plan as inputs that are processed to output a likely range of environmental state changes. The configuration of the Bayesian network is performed by one or more research scientists and / or data analysts 28 using known techniques for assigning probabilities within the Bayesian network.

[0041] If initially configured solely or primarily based on expert opinion, the output of the Bayesian network is likely to have a high degree of uncertainty. That is, the range of environmental condition changes predicted by the Bayesian network is likely to be relatively wide. For this reason, a cautious approach should be taken to the initial operation of the extraction system. This may require the use of conservative extraction plans early in the project lifecycle, and initial extraction activities may need to be located first in areas deemed to have greater environmental resilience. The purpose of the system is to assist key project personnel in promoting a precautionary approach that emphasizes proactive decision-making to manage risk. This approach also encourages delaying potentially harmful decisions until a sufficiently detailed understanding of the relevant causal relationships is obtained. As shown in Figure 4, using active adaptive management techniques allows the extraction system to safely explore operating states where data is absent or minimal. Once a historical record of empirically derived data is compiled from completing and monitoring actual extraction activities, this empirically derived data is used by one or more research scientists and / or data analysts 28 to refine the configuration of the Bayesian network. Each time this reconfiguration is performed, predictions from the probabilistic analysis will typically have lower uncertainty. That is, the range of environmental state changes predicted by the Bayesian network is likely to become narrower for a given confidence interval. Once this state is reached, the initial cautious approach can be gradually replaced by a more aggressive approach, to the extent justified by the certainty of the probabilistic analysis' predictions. Once the operational state space has been fully explored and the user deems it appropriate, active adaptive management techniques can be replaced by more traditional passive adaptive management techniques, as shown in Figure 3.

[0042] Computing system 1 maintains a mode indicator that indicates whether the system is currently desired to operate in an active adaptive management state or a passive adaptive management state. It will be appreciated that passive adaptive management, such as that shown in Figure 3, executes a single control strategy before collecting new data and re-evaluating whether that control strategy was effective. In contrast, active adaptive management, as shown in Figure 4, uses models and data derived from the implementation of harvesting activities to compare multiple control strategies to a control.

[0043] In one embodiment, the computer system switches the mode indicator between an active adaptive management state and a passive adaptive management state in response to user input. In another embodiment, the computer system 1 is configured to calculate an uncertainty score associated with the probabilistic analysis, and when the uncertainty score falls below a threshold, the computer system recommends to the user to set the mode indicator to the active adaptive management state. When the uncertainty score rises above the threshold, the computer system automatically sets the mode indicator to the passive adaptive management state. Preventing the computing system 1 from automatically setting the mode indicator to the active adaptive management state is a security feature because the mode cannot be changed to the active adaptive management state without the knowledge and consent of the system's user.

[0044] To begin the process shown in FIG. 4 , a user of computer system 1, typically a production manager, establishes an initial proposed harvest plan for the extraction of natural resources from the target environment. In practice, multiple harvest plans may be implemented simultaneously at different sites within the target environment; the example shown in FIG. 4 involves two harvest plans being implemented simultaneously. However, it will be understood that other embodiments may implement fewer than two or more harvest plans simultaneously. The establishment of one or more initial harvest plans can be done at several possible levels of specificity. At the highest level, this may simply involve defining the proposed production goals and proposed area of ​​the target environment where extraction will occur. At lower levels, it may involve defining more operational details of the proposed harvesting activities. Optionally, the initial proposed harvest plan may also include other desired parameters, such as efficiency, power consumption, etc., which are entered into computing system 1 and stored as plan data for further processing. In the example shown in FIG. 4 , in step 4.1, a production manager enters plan data for two harvest plans proposed to be implemented simultaneously.

[0045] The data representing the two initial proposed harvest plans is subject to a "sanity check," which involves processing by a physics simulation module of computing system 1 to check whether the two initial proposed harvest plans can be physically implemented. If not, the user is prompted to revise the initial proposed harvest plan(s).

[0046] Assuming the "sanity check" passes, the data representing the two initial proposed collection plans is processed by an optimization algorithm, such as a multidisciplinary design optimization (MDO) engine. The MDO engine is configured to process the plan data and environmental logic 8 to generate specific details for each of the two initial proposed collection plans that meet the specified initial criteria. This includes productivity targets, resource collection rates, detailed proposed trajectories with associated timestamps for the collector 4, riser 5, and any mobile operating equipment, such as the vessel 6 or ROV 7, schedules of operations, etc., as well as proposed riser flow rates and other relevant operational data. In some embodiments, as shown in steps 6.1 through 6.3 of FIG. 6 , an initial compliance check is performed, including simulated processing of the environmental, operational, and plan data for each of the two initial proposed collection plans, to determine whether any of the environmental constraints 10 are simulated to be violated by either of the initial proposed collection plans. If the simulation predicts that any of the initial proposed harvest plans is likely to violate any of the environmental constraints 10, computing system 1 is configured to flag the corresponding harvest plan as invalid in step 6.3 and prompt the user to revise the corresponding harvest plan in step 6.4 to generate an alternative harvest plan for operational implementation. This may include the user selecting at least one of a number of predetermined potential revisions to the harvest plan, such as, for example, reducing production targets by 10%, or shifting the proposed harvest site to a less sensitive area or away from a protected location.

[0047] When a tracked ecosystem variable is predicted to approach, but not exceed, a tolerance threshold, increased monitoring is triggered, improving the system's ability to more accurately predict whether an at-risk indicator will violate a constraint.

[0048] Once the two initial proposed harvest plans pass the above checks and are approved by the necessary personnel, they are provided to operational staff in steps 4.4A and 4.4B to begin parallel implementation of the two harvest plans within the target environment. At the start of the parallel implementation, computing system 1 starts a timer that serves as a starting point for three time periods. The first time period is the period over which the simulation will run. In a typical implementation, the first time period is likely to be between about one month and about one year.

[0049] The second period is a period during which the two initial harvesting plans are implemented in parallel within the target environment and is shorter than the first and third periods. Typically, the second period is likely to be between about five days and about six months. During this second period, multiple alternative harvesting plans are generated and simulated, as described in more detail below.

[0050] Having a longer first period compared to the second period means that techniques from model predictive control can be implemented. That is, the simulated period covered by the simulation process is longer than the period over which the initial two harvest plans are proposed to be operationally implemented. This longer period advantageously allows for a more thorough calculation of the environmental impacts of various harvest plans. Furthermore, if the generation of multiple alternative compliant harvest plans were to fail by the end of the second period, the longer period allows production managers to have confidence in deciding whether to continue implementing the two initial harvest plans. This is because the environmental impacts have already been predicted for a long period beyond the end of the second period.

[0051] As shown in FIG. 4, six alternative harvest plans are generated in step 4.3. This generation process is shown in more detail in FIG. 7. This process uses the optimization algorithm described above. However, rather than receiving optimization parameters from a user, in this case, computer system 1 is configured to automatically generate alternative starting parameters for optimization in step 7.1. Two of the six alternative harvest plans will be selected to be implemented at the end of the predetermined second time period. Thus, the starting operating parameters of the first set of three alternative harvest plans 4.5A, 4.5B, and 4.5C correspond to the predicted operating parameters of one of the implemented harvest plans 4.4A as the predetermined second time period elapses. This means that the states of the operating equipment of the implemented harvest plan 4.4A at the end of the predetermined second time period will match the simulated starting states of the operating equipment of the three alternative harvest plans 4.5A, 4.5B, and 4.5C. Similarly, the starting operating parameters of the second set of three alternative harvest plans 4.5D, 4.5E, and 4.5F correspond to the predicted operating parameters of the other of the implemented harvest plans 4.4B as the predetermined second period elapses, meaning that the states of the operating equipment of the implemented harvest plan 4.4B at the end of the predetermined second period will match the simulated starting states of the operating equipment of the three alternative harvest plans 4.5D, 4.5E, and 4.5F.

[0052] One strategy for generating alternative starting parameters for processing by the MDO engine to generate alternative harvesting plans is for the computing system 1 to be configured to incrementally modify the starting parameters (e.g., some manipulated variables) of the two user-generated harvesting plans. Examples of such manipulated variables that may be incrementally modified include the proposed harvesting start location within the target environment, the proposed harvesting rate of the natural resource, the proposed efficiency level, the proposed power usage level, the waypoints (x, y, depth) of each collector's trajectory, the collection rate associated with each waypoint, the speed of the collector 4, etc. Another strategy for generating alternative starting parameters for optimization is for the computing system 1 to be configured to randomly generate the harvesting plan starting parameters. This random element helps avoid any local minima that may be associated with the user-generated harvesting plans. However, it will be understood that not all starting parameters can be changed, and that randomly generated starting parameters must be simulated in a physics simulator to ensure that they are achievable within the laws of physics. Typically, both of the above strategies will be used to generate multiple sets of alternative optimization starting parameters in step 7.1.

[0053] In step 7.2, the computing system 1 adjusts the optimization algorithm depending on the current status of the mode indicator. If the mode indicator is in a passive adaptive management state, the optimization algorithm is adjusted to ensure that the generation of multiple alternative harvest plans is constrained within a safe region of the operational envelope. In other words, the alternative harvest plans generated by using the MDO engine to optimize based on a set of alternative optimization starting parameters are likely to comply with environmental constraints.

[0054] When the mode indicator is in the active adaptive management state, the optimization algorithm is adjusted to ensure that the generation of multiple alternative harvest plans is biased toward the boundary region of the operating envelope. In other words, alternative harvest plans generated by using the MDO engine to optimize based on alternative sets of optimization starting parameters will either be nearly non-compliant with the environmental constraints 10 or not compliant at all with the environmental constraints 10. In one embodiment, computing system 1 maintains a variable that defines the degree of this bias. Upon completion of the MDO engine optimization, multiple alternative harvest plans and their respective plan data will have been generated.

[0055] Each of the six alternative harvest plans is then subjected to a simulation process in steps 4.5A, 4.5B, 4.5C, 4.5D, 4.5E, and 4.5F (see also step 7.3 in FIG. 7 ). This example shows a total of six alternative harvest plans. However, in practice, as many alternative harvest plans as computationally feasible would typically be generated, simulated, and compliance-checked within the second time period. The simulation process processes the environmental data, operational data, and planning data for each of the six alternative harvest plans to simulate the respective state of the target environment after a predetermined first time period has passed during which each of the alternative harvest plans is simulated to be implemented. In other words, for each of the alternative harvest plans, the simulation process determines a range of predicted end states of the target environment after each of the alternative harvest plans has been simulated to be implemented for a length of time equal to the first time period.

[0056] The simulation process takes as input the time history of operational and environmental data, along with planning data for alternative harvest plans generated by the MDO engine optimization, and uses them to run future simulations. The simulations are run into the future to cover a first predetermined time period using a world simulator 19 and a collection of subsystem simulators 20, such as a materials processing simulator, a biological simulator, and a physics simulator. For example, the physics simulators are used by the computing system 1 in conjunction with ocean current forecast data to simulate the dispersion and settling of a sediment plume.

[0057] A probabilistic analysis using a Bayesian network is performed using the final state of the simulation to determine the most likely range of values ​​for the key ecosystem indicators for each of the alternative extraction plans at the end of a predetermined first period. In the deep sea mining example, the key ecosystem indicators include: primary production · Surface photosynthesis Phytoplankton density / biomass Nutrient concentration · Chemical synthesis o Water chemistry Carbon flux o Plankton community composition Bioturbation o Sediment characteristics o Caste density Precipitation Radiochemistry Biodiversity · Habitat integrity o Plume deposition (benthic) o Integrity of the conservation area o Plume dispersion (mid-deep sea) Fauna characterization o Number of major species Noise generation Microbial diversity o Number of major species Nutritional support o eDNA analysis

[0058] A 90% confidence interval is used to determine the likely range of values ​​for each ecosystem indicator for each alternative harvest plan. If any part of the range of possible values ​​falls outside the acceptable range, the alternative harvest plan is flagged as non-compliant.

[0059] The output of the simulation process allows for the calculation of the degree of human impact resulting from the simulated harvesting activities of each of the six alternative harvesting plans by subtracting the final simulated environmental state from the initial simulated environmental state. Constraint processing is then performed, whereby the constraint data and simulated states of the target environment for each of the six alternative harvesting plans are processed to determine whether any of the environmental constraints are simulated to be violated. If none of the environmental constraints are simulated to be violated for one of the alternative harvesting plans, computing system 1 is configured to flag the harvesting plan as compliant. If at least one of the environmental constraints is simulated to be violated for one of the alternative harvesting plans, computing system 1 is configured to flag the harvesting plan as non-compliant. Thus, the objective is to compile a plurality of compliant simulated alternative harvesting plans by the end of the second time period.

[0060] The above simulation and constraint processing occurs in parallel with the operational implementation of the two harvesting plans during a predetermined second time period. During this implementation, computing system 1 continues to monitor data coming from sensors 3, as shown in step 4.6. This empirically derived environmental data and empirically derived operational data are stored by computing system 1 for later use (see also step 6.8 of FIG. 6 and step 8.1 of FIG. 8).

[0061] At the end of the second period, computing system 1 is configured to calculate a metric for each of the six alternative harvest plans. This metric is formulated to quantify the desirability of the harvest plan for which it is calculated. It is typically calculated by taking a weighted average of various criteria. The three metrics for the first set of alternative harvest plans 4.5A, 4.5B, and 4.5C are compared, and the alternative harvest plan with the highest metric is then presented to the production manager for approval to be implemented (see also step 6.5 of Figure 6). Similarly, the three metrics for the second set of alternative harvest plans 4.5D, 4.5E, and 4.5F are compared, and the alternative harvest plan with the highest metric is also presented to the production manager for approval to be implemented. Once approved, the process flow loops through inner loop 4.7 and resumes with the implementation of the two newly approved harvest plans (also shown in steps 6.6 and 6.7 of Figure 6).

[0062] Inner loop 4.7 continues looping as described above until a third period of time has elapsed. In a typical implementation, the third period of time is likely to be between about one month and about one year. Once the third period of time has elapsed, computing system 1 is configured to reconcile the previously stored environmental data with the empirically derived operational data during the inner loop, as shown in FIG. 4 (see also step 8.2 of FIG. 8). In step 4.8 (see also step 6.9 of FIG. 6 and step 8.3 of FIG. 8), one or more research scientists and / or data analysts 28 apply probabilistic and / or machine learning techniques to the empirically derived environmental data and the empirically derived operational data to update any of the probabilistic analysis, constraint data 10, tolerance levels for state changes 12, and / or ecosystem model 9. Process flow returns to step 4.2 via outer loop 4.9, where environmental logic 8 is updated (see also step 8.4 of FIG. 8). Thus, environmental logic 8 no longer relies on initial expert input. Rather, they are updated with reference to empirically derived data, and therefore future predictions resulting from simulations using environmental logic 8 are likely to benefit from improved certainty. This improved certainty gives computing system 1 the freedom to generate alternative harvesting plans that safely explore closer to the outer edge of the operational envelope when subsequently executing steps 4.3 through 4.7 of the inner loop.

[0063] As best shown in FIG. 1 , computing system 1 is configured to maintain a portal 13 accessible to regulators 14. Computing system 1 also maintains a portal 15 accessible to the public. Users of computing system 1 can tailor these portals 13, 15 to select the information they make available. This selection is made from among environmental data, constraint data, and operational data. Typically, the portal 13 accessible to regulators 14 is likely to include more detailed information compared to the information made available via the public portal 15. Computing system 1 is also configured to maintain a pair of dashboards 23 and 26 that summarize key information, primarily retrieved from digital twin 2, likely to be needed by harvesting equipment operations managers 25 and operators 27, respectively.

[0064] Embodiments of the present invention take environmental and production data as inputs in near real time and output harvest plans, strategies, mitigations, actions, data visualizations, and controls for execution by human operators. This helps facilitate environmental management strategies that adapt to current operating conditions and predicted future states, and are also sensitive to the system's statistical confidence in the cause-and-effect relationships defined in its ecosystem model. This approach involves continuous testing of hypotheses, collecting data, and updating environmental parameter values ​​to enable operations to adapt to the current and predicted future states of the affected environment. Embodiments of the present invention enable key personnel to modify the location, approach, and specifications of the harvest plan based on changes in environmental factors, and these operational changes can be implemented within weeks rather than months or years. Importantly, embodiments of the present invention enable dynamic changes to the harvest plan to be made during operations. This is an advantage over prior art, where such changes could typically be made only before the start of a harvesting activity.

[0065] While certain preferred embodiments have been described, it will be appreciated by those skilled in the art that numerous variations and / or modifications may be made to the present invention without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

Claims

1. 1. A computer-implemented process for processing a harvest plan for harvesting natural resources from a target environment, the process comprising: accessing environmental data indicative of the target environment; access constraint data indicative of a plurality of environmental constraints applicable to the target environment; having access to operational data indicative of devices deployed within the target environment; accessing planning data indicative of the collection plan; executing a simulation process to process the environmental data, the operational data, and the plan data to simulate conditions in the target environment after a predetermined first period of time during which the harvesting plan is simulated to be implemented; performing constraint processing that processes the constraint data and the simulated state of the target environment to determine whether any of the environmental constraints are simulated to be violated, and flagging the harvest plan as compliant if none of the environmental constraints are simulated to be violated; generating a plurality of alternative harvest plans and respective plan data and performing a simulation and constraint process on each of the respective plan data to flag a plurality of conforming harvest plans; A computer-implemented process comprising:

2. 2. The computer-implemented process of claim 1, comprising: defining metrics for quantifying desirability of collection plans; calculating individual metrics for each of the compliant collection plans; and presenting the collection plans with the highest metrics to a user of the computing system for approval after a second predetermined period of time has elapsed, the second predetermined period of time being shorter than the first predetermined period of time.

3. 10. The computer-implemented process of claim 1, wherein if an environmental constraint is simulated as being violated with respect to the harvest plan, the computing system is configured to prompt a user of the computing system to select at least one of a plurality of predetermined potential revisions to the harvest plan.

4. 10. The computer-implemented process of claim 1, wherein the simulation process uses a digital twin configured to simulate current operational conditions based on operational data input from sensors, and the digital twin is configurable to simulate future operational data in which a harvesting plan is simulated to be implemented.

5. 5. The computer-implemented process of claim 4, wherein the simulation process uses probabilistic analysis to process the operational data and the future operational data to model causal relationships for multiple indicators of environmental impact.

6. 2. The computer-implemented process of claim 1, wherein generating the plurality of alternative harvest plans and respective plan data comprises incrementally modifying harvest plan manipulated variables to generate modified manipulated variables and processing the modified manipulated variables with an optimization algorithm.

7. 2. The computer-implemented process of claim 1, wherein generating the plurality of alternative harvest plans and respective planning data comprises randomly generating harvest plan initiation parameters and subjecting the randomly generated initiation parameters to an optimization algorithm.

8. 8. The computer-implemented process of claim 7, wherein the starting parameters include at least one of a proposed harvesting start location within the target environment, a proposed harvesting rate of natural resources, a proposed efficiency level, and a proposed power usage level.

9. The computer-implemented process of claim 1 , wherein the computing system maintains a mode indicator that indicates an active adaptive management state or a passive adaptive management state.

10. 10. The computer-implemented process of claim 9, wherein when the mode indicator is in the passive adaptive management state, generating the plurality of alternative harvest plans is constrained within a safety region of an operating envelope, whereby the alternative harvest plans are likely to comply with the environmental constraints.

11. 10. The computer-implemented process of claim 9, wherein when the mode indicator is in the active adaptive management state, generating the plurality of alternative harvest plans biases toward boundary regions of an operating envelope where the alternative harvest plans are nearly non-compliant with or non-compliant with the environmental constraints.

12. The computer-implemented process of claim 5 , wherein the computer system switches the mode indicator between an active adaptive management state and a passive adaptive management state in response to user input.

13. 6. The computer-implemented process of claim 5, wherein the computer system is configured to calculate an uncertainty score associated with the probabilistic analysis, and when the uncertainty score falls below a threshold, the computer system recommends to a user to set the mode indicator to an active adaptive management state, and when the uncertainty score rises above the threshold, the computer system automatically sets the mode indicator to a passive adaptive management state.

14. 10. The computer-implemented process of claim 1, further comprising: operatively implementing a collection plan during a predetermined second time period; and generating the plurality of alternative collection plans and respective plan data during the predetermined second time period.

15. 15. The computer-implemented process of claim 14, wherein starting operating parameters of at least some of the alternative harvest schedules correspond to projected operating parameters of the implemented harvest schedule at the elapse of the second predetermined period of time.

16. 2. The computer-implemented process of claim 1, further comprising: defining a third time period, the third time period being longer than the second time period, and wherein, once the third time period has elapsed, the computing system is configured to match empirically derived environmental data with empirically derived operational data.

17. Research scientists and / or data analysts: said probabilistic analysis; the constraint data; Tolerance level of condition change, and / or Ecosystem models, 6. The computer-implemented process of claim 5, further comprising applying probabilistic and / or machine learning techniques to the empirically derived environmental data and the empirically derived operational data to update at least one of:

18. 17. The computer-implemented process of claim 16, wherein the second period of time is between five days and six months, and the first and third periods of time are each between one month and one year.

19. 10. The computer-implemented process of claim 1, wherein the computing system is configured to maintain a portal accessible to regulators and / or the public, the portal making available at least one type of information: environmental data, constraint data, and operational data.

20. A computing system configured to perform the process of claim 1.

21. 1. A system for harvesting natural resources from a target environment, comprising: A computing system configured to perform the process of claim 1; a collection device deployed in the target environment; a plurality of sensors deployed on and around the sampling device and within the target environment, the plurality of sensors communicating with the computing system via a telemetry link; A system comprising:

Images