US-12626061-B1 - Converting a process industry standard operating procedure into a simulatable non-ambiguous single state machine architecture

US12626061B1US 12626061 B1US12626061 B1US 12626061B1US-12626061-B1

Abstract

Techniques are described for automatically converting a standard operating procedure (SOP) into a simulatable non-ambiguous single state machine (SSM) architecture in which a given input (transition condition) necessarily and unambiguously moves the system from a first state to a predictable second state. By converting a written SOP into a simulatable non-ambiguous SSM architecture, the techniques described herein reduce the possibility of errors and improve the safety, efficiency, and operational performance of the system. The described techniques also allow the architecture, and indeed the entire system, to be simulated.

Inventors

Jaime Aguilera
Juan Bautista Gomez
Fernando Alonso
Jon Zamora
Christian McDermott
Francis Montemurro

Assignees

Voovio Technologies SL

Dates

Publication Date: 20260512
Application Date: 20220214

Claims (20)

1 . A computer-implemented method for converting a standard operating procedure defining a process into a non-ambiguous single state machine architecture, comprising: at an input device, receiving as input a standard operating procedure comprising a series of instructions, wherein each instruction specifies an action; and for each action specified by an instruction in the standard operating procedure: at a hardware processor, automatically determining whether the instruction specifies the action in an ambiguous manner; at the hardware processor, responsive to the instruction being specified in an ambiguous manner: obtaining a description of the ambiguity; and removing the ambiguity from the instruction; at the input device, obtaining a description of the action to be performed; at the hardware processor, based on the obtained description, automatically generating a predefined input for the single state machine architecture; and at an electronic storage device, storing the predefined input for the single state machine architecture; wherein automatically determining whether the instruction specifies the action in an ambiguous manner comprises: obtaining text comprising words specifying the action; tokenizing the text; applying a language model to the tokenized text; tagging parts of speech in the tokenized text, to generate a plurality of tags; performing syntactic parsing on at least one of the tags and the words in the text; and applying a machine-learning based ambiguity routine to the output of the syntactic parsing step, to automatically determine whether the action is specified in an ambiguous manner.
2 . The method of claim 1 , wherein obtaining a description of the action to be performed comprises: receiving input from an individual capturing information about the action by observation.
3 . The method of claim 2 , wherein the individual is a subject matter expert.
4 . The method of claim 2 , wherein the instruction specifies an action to be performed on a piece of equipment, and wherein receiving input from the individual capturing information about the action comprises receiving input from the individual capturing information about the action by observation of the equipment to be acted upon.
5 . The method of claim 4 , wherein receiving input from the individual capturing information about the action comprises receiving at least one selected from the group consisting of: a field location of the equipment to be acted upon; at least one image of the equipment; at least one image of the environment surrounding the equipment; at least one verb specifying the action to be performed on the equipment; a name of the equipment; and a reference to a step of the standard operating procedure.
6 . The method of claim 1 , wherein automatically determining whether the instruction specifies the action in an ambiguous manner further comprises at least one selected from the group consisting of: automatically determining whether all terminology used in the instruction is unequivocal; automatically determining whether the action specified by the instruction references a system or a single component; automatically determining whether the action specified by the instruction specifies a single object or an indefinite number of objects; automatically determining whether performing the action requires any previous operational know-how.
7 . The method of claim 1 , wherein the process defined by the standard operating procedure comprises a process for transforming a first material into a second material.
8 . The computer-implemented method of claim 1 , wherein the language model comprises a Unigram Language Model.
9 . A computer-implemented method for converting a standard operating procedure defining a process into a non-ambiguous single state machine architecture, comprising: at an input device, receiving as input a standard operating procedure comprising a series of instructions, wherein each instruction specifies an action; and for each action specified by an instruction in the standard operating procedure: at a hardware processor, determining whether the instruction specifies the action in an ambiguous manner; at the hardware processor, responsive to the instruction being specified in an ambiguous manner: obtaining a description of the ambiguity; and removing the ambiguity from the instruction; at the input device, obtaining a description of the action to be performed; at the hardware processor, based on the obtained description, automatically generating a predefined input for the single state machine architecture; and at an electronic storage device, storing the predefined input for the single state machine architecture; wherein the process defined by the standard operating procedure comprises a process for transforming a first material into a second material by performing at least one selected from the group consisting of: applying pressure; applying temperature; and adding additional materials.
10 . A non-transitory computer-readable medium for converting a standard operating procedure defining a process into a nonambiguous single state machine architecture, comprising instructions stored thereon, that when performed by a hardware processor, perform the steps of: causing an input device to receive as input a standard operating procedure comprising a series of instructions, wherein each instruction specifies an action; and for each action specified by an instruction in the standard operating procedure, performing the steps of: automatically determining whether the instruction specifies the action in an ambiguous manner; responsive to the instruction being specified in an ambiguous manner: obtaining a description of the ambiguity; and removing the ambiguity from the instruction; causing the input device to obtain a description of the action to be performed; based on the obtained description, automatically generating a predefined input for the single state machine architecture; and causing an electronic storage device to store the predefined input for the single state machine architecture; wherein automatically determining whether the instruction specifies the action in an ambiguous manner comprises: obtaining text comprising words specifying the action; tokenizing the text; applying a language model to the tokenized text; tagging parts of speech in the tokenized text, to generate a plurality of tags; performing syntactic parsing on at least one of the tags and the words in the text; and applying a machine-learning based ambiguity routine to the output of the syntactic parsing step, to automatically determine whether the action is specified in an ambiguous manner.
11 . The non-transitory computer-readable medium of claim 10 , wherein causing the input device to obtain a description of the action to be performed comprises: causing the input device to receive input from an individual capturing information about the action by observation.
12 . The non-transitory computer-readable medium of claim 11 , wherein the individual is a subject matter expert.
13 . The non-transitory computer-readable medium of claim 11 , wherein the instruction specifies an action to be performed on a piece of equipment, and wherein causing the input device to receive input from the individual capturing information about the action comprises causing the input device to receive input from the individual capturing information about the action by observation of the equipment to be acted upon.
14 . The non-transitory computer-readable medium of claim 13 , wherein causing the input device to receive input from the individual capturing information about the action comprises causing the input device to receive at least one selected from the group consisting of: a field location of the equipment to be acted upon; at least one image of the equipment; at least one image of the environment surrounding the equipment; at least one verb specifying the action to be performed on the equipment; a name of the equipment; and a reference to a step of the standard operating procedure.
15 . The non-transitory computer-readable medium of claim 10 , wherein determining whether the instruction specifies the action in an ambiguous manner further comprises at least one selected from the group consisting of: automatically determining whether all terminology used in the instruction is unequivocal; automatically determining whether the action specified by the instruction references a system or a single component; automatically determining whether the action specified by the instruction specifies a single object or an indefinite number of objects; automatically determining whether performing the action requires any previous operational know-how.
16 . The non-transitory computer-readable medium of claim 10 , wherein the process defined by the standard operating procedure comprises a process for transforming a first material into a second material.
17 . The non-transitory computer-readable medium of claim 10 , wherein the language model comprises a Unigram Language Model.
18 . A non-transitory computer-readable medium for converting a standard operating procedure defining a process into a nonambiguous single state machine architecture, comprising instructions stored thereon, that when performed by a hardware processor, perform the steps of: causing an input device to receive as input a standard operating procedure comprising a series of instructions, wherein each instruction specifies an action; and for each action specified by an instruction in the standard operating procedure, performing the steps of: determining whether the instruction specifies the action in an ambiguous manner; responsive to the instruction being specified in an ambiguous manner: obtaining a description of the ambiguity; and removing the ambiguity from the instruction; causing the input device to obtain a description of the action to be performed; based on the obtained description, automatically generating a predefined input for the single state machine architecture; and causing an electronic storage device to store the predefined input for the single state machine architecture; wherein the process defined by the standard operating procedure comprises a process for transforming a first material into a second material by performing at least one selected from the group consisting of: applying pressure; applying temperature; and adding additional materials.
19 . A system for converting a standard operating procedure defining a process into a non-ambiguous single state machine architecture, comprising: an input device, configured to receive as input a standard operating procedure comprising a series of instructions, wherein each instruction specifies an action; a hardware processor, communicatively coupled to the input device, configured to, for each action specified by an instruction in the standard operating procedure: automatically determine whether the instruction specifies the action in an ambiguous manner; responsive to the instruction being specified in an ambiguous manner: obtain a description of the ambiguity; and remove the ambiguity from the instruction; cause the input device to obtain a description of the action to be performed; and based on the obtained description, automatically generate a predefined input for the single state machine architecture; and an electronic storage device, communicatively coupled to the hardware processor, configured to store each generated predefined input for the single state machine architecture; wherein automatically determining whether the instruction specifies the action in an ambiguous manner comprises: obtaining text comprising words specifying the action; tokenizing the text; applying a language model to the tokenized text; tagging parts of speech in the tokenized text, to generate a plurality of tags; performing syntactic parsing on at least one of the tags and the words in the text; and applying a machine-learning based ambiguity routine to the output of the syntactic parsing step, to automatically determine whether the action is specified in an ambiguous manner.
20 . The system of claim 19 , wherein causing the input device to obtain a description of the action to be performed comprises: causing the input device to receive input from an individual capturing information about the action by observation.

Description

CROSS-REFERENCE TO RELATED APPLICATION The present application claims the benefit of U.S. Provisional Application Ser. No. 63/149,893 for “Converting a Process Industry Standard Operating Procedure into a Simulatable Non-Ambiguous Single State Machine Architecture”, filed on Feb. 16, 2021, which is incorporated by reference herein in its entirety. TECHNICAL FIELD The present document relates to techniques for generating a single state machine architecture to unambiguously define a procedure. BACKGROUND A standard operating procedure (SOP) is a set of step-by-step instructions compiled by an organization to help operators carry out routine and non-routine operations. SOPs aim to achieve safety, efficiency, quality output, and uniformity of performance, while reducing miscommunication and failure to comply with industry regulations. SOPs are often used in the manufacturing industry, as a way to instruct operators as to proper equipment and process operations. Conventionally, SOPs are often created from a piping and instrumentation diagram (P&ID) or similar diagram describing a series of steps to be performed. See, for example, Landon et al., U.S. Pat. No. 9,613,233, for Interactive Industrial Maintenance, Testing, and Operation Procedures. Such conventional methods suffer from a number of deficiencies. The SOPs are a product of the process technology and are usually written by technical consultants or subject matter experts, and often rely on knowledge and expertise of a select few. As a result, the SOPs often include high-level generalizations, inconsistencies, and/or ambiguities, and fail to provide effective checks to prevent actions from being executed incorrectly. In general, SOPs do not include sufficient supporting guidance or information, such as images that might help illustrate details of the procedure and support mapping and understanding of the procedure to a real-world situation. This lack of procedure clarity results in operating practices that are not consistent with established SOPs. SUMMARY In various embodiments, the system and method described herein provide improved efficiency and reduced ambiguity by converting an SOP into a simulatable non-ambiguous single state machine (SSM) architecture. An example is shown in FIG. 1. In this manner, a given input (transition condition) necessarily and unambiguously moves the system from a first state (designated State A) to a predictable second state (designated State B). In various embodiments, the system and method described herein are able to convert a written SOP into a simulatable non-ambiguous SSM architecture, thus reducing the possibility of errors and improving the safety, efficiency, and operational performance of the system. The described techniques also allow the architecture, and indeed the entire system, to be simulated. More specifically, a SSM architecture can be simulated with a logic engine that is able to read and change the value of all the variables in the machine. In each state, the value of the variables specified, and when a transition condition is met (such as, for example, changing a value of a variable from 0 to 1), the system updates the variable values of the entire architecture. Further details and variations are described herein. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, together with the description provided below, illustrate several embodiments. One skilled in the art will recognize that the particular embodiments illustrated in the drawings and described herein are merely exemplary, and are not intended to limit scope. FIG. 1 is a block diagram depicting a hardware architecture for implementing the techniques described herein according to one embodiment. FIG. 2 is a block diagram depicting a hardware architecture for implementing the techniques described herein in a client/server environment, according to one embodiment. FIG. 3 is a block diagram depicting an example of a single state machine (SSM). FIG. 4 depicts an example of an SOP as may be used in a process industry. FIG. 5 is a flow diagram depicting a method for converting an SOP to an SSM while ensuring that there are no inconsistencies or ambiguities, according to one embodiment. FIG. 6 is a flow diagram depicting additional details for checking for and resolving ambiguities, according to one embodiment. FIG. 7 is a table depicting examples of methods for rewriting actions for an SOP, according to one embodiment. FIG. 8 is a flow diagram depicting a method by which a subject matter expert (SME) can investigate the locations of components to determine whether ambiguity exists, according to one embodiment. FIG. 9 is a block diagram depicting a methodology for capturing information in the field to generate a non-ambiguous instruction, or to rewrite an ambiguous instruction as a non-ambiguous one, according to one embodiment. FIGS. 10A and 10B are screen shots depicting an example of an application interface for executing the methodology descri