Injury risk factor identification, prediction, and mitigation

• US 10,453,015 B2
• Filed: 07/29/2015
• Issued: 10/22/2019
• Est. Priority Date: 07/29/2015
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Chandra, Nilanjana, Goyal, Munish, Gridley, Anthony, Squires, Brett A., Ye, LanXiang
• Primary Examiner(s): Waldron, Scott A.
• Assistant Examiner(s): Lamardo, Viker A
• Application Number:
  US14/811,950
• Publication Number:
  US 20170032255A1
• Time in Patent Office: 1,546 Days
• Field of Search: 706 11
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities

Abstract

A method and system for identifying workplace risk factors is provided. The method includes monitoring via execution of multiple geographically distributed sensor devices, workplace injury based events associated with individuals at a multisite distributed workplace environment. Current injury data describing the workplace injury based events is stored and predicted future workplace injury based events associated with future workplace injury based events with respect to a predicted plurality of individuals at the multisite distributed workplace environment are determined. Injury risk mitigating actions associated with prevention of said predicted future workplace injury based events are generated and an associated cost optimized reduction plan for prioritized implementation of the injury risk mitigating actions is generated.

Claims

1. A workplace risk factor identification and safety issue incident resolution method comprising:

monitoring, by a computer processor of an embedded controller comprising internal dedicated electronic circuitry comprising a programmable logic array utilizing state information of specialized firmware program code configured to execute specialized software for personalizing the electronic circuitry executing multiple geographically distributed dedicated hardware sensor devices, a plurality of workplace injury based events associated with a plurality of individuals at a multisite distributed workplace environment, wherein said dedicated hardware sensor devices comprise RFID sensors for identification, video camera based sensors for automatically recording said plurality of workplace injury based events, and mobile device type sensors for near real time reporting of said plurality of workplace injury based events.,

tagging, by said computer processor, said plurality of workplace injury based events;

storing, by said computer processor within a centralized server, current injury data describing said plurality of workplace injury based events comprising tags from said tagging;

presenting, by said computer processor via a first clickable interactive view portion of a first GUI interface in response to activating said first clickable interactive view portion, a frequency of injury type at selected locations;

presenting, by said computer processor via a second clickable interactive view portion of second GUI interface in response to activating said second clickable interactive view portion, injury locations associated with specified zip codes;

presenting, by said computer processor via a third clickable interactive view portion of a third GUI interface in response to activating said clickable interactive view portion, repair actions associated with said plurality of workplace injury based events;

performing, by said computer processor, a hierarchical clustering process with respect to incident keywords to obtain a set of distinct incident types defined by incident description text;

executing, by said computer processor, a set of time series models for each incident cluster type to predict a rate of future occurrence for each incident type of said set of distinct incident types;

executing, by said computer processor, a clustering algorithm to cluster demographic attributes by occurrence of frequencies across all identified incident types;

extracting, by said computer processor, demographic profiles of an active population of said plurality of individuals;

predicting, by said computer processor, a cluster id for an associated demographic profile of said demographic profiles;

normalizing, by said computer processor, each cluster level prediction by evaluating a proportional contribution of a demographic feature vector with respect to each said incident type;

combining, by said computer processor, said set of time series models to predict an incident type, a rate, and a detailed analytical model with respect to employee demographic features;

generating, by said computer processor based on said current injury data, said frequency of injury type at selected locations, said injury locations, said repair actions, and previously retrieved injury data describing previously retrieved historical workplace injury based events, predicted future workplace injury based events associated with future workplace injury based events with respect to a predicted plurality of individuals at said multisite distributed workplace environment;

generating, by said computer processor based on said predicted future workplace injury based events, injury risk mitigating actions associated with prevention of said predicted future workplace injury based events;

identifying, by said computer processor in response to said executing said clustering algorithm and via a text analytics process applied to an injury log, root causes for said plurality of workplace injury events with respect to injury signatures of pervious work related injuries to identify distinct injury type signatures caused by activities monitored via said RFID sensors, said video camera based sensors, and said mobile device type sensors;

evaluating, by said computer processor, probabilistic map with respect to an injury signature of said distinct injury type signatures associated with a demographic pattern via a series of classification process;

identifying, by said computer processor, a root cause, of said root causes, with respect to each said injury signature via a text analytics process applied to said injury log;

executing said injury risk mitigating actions such that said root causes are corrected thereby mitigating injury risks associated with workplace safety resulting in a maximal gain in productivity, incident risk resolution options, and reduced injury rates with respect to a frequency, severity, and expected loss associated with body stress events, motor vehicle accident related events, bodily injury related events, medical condition related events, biological factor related events, mental stress related events, chemical accident related events, and heat related events;

generating, by said computer processor based said injury risk mitigating actions, a cost optimized reduction plan for prioritized implementation of said injury risk mitigating actions; and

deploying, maintaining, and integrating computer-readable code into said embedded controller for execution of said method associated with improving said incident risk resolution options and reducing workplace injury rates.

2. The method of claim 1, further comprising:

transmitting, by said computer processor to a mobile device executing an interactive spatio-temporal application, said injury risk mitigating actions and said cost optimized reduction plan for presentation to a user.

3. The method of claim 1, wherein said predicted future workplace injury based events are generated with respect to future times of occurrence, future locations and frequencies of occurrence, and demographics with respect to said predicted plurality of individuals at said multisite distributed workplace environment.

4. The method of claim 1, wherein said generating said predicted future workplace injury based events comprises:

identifying event types associated with said predicted future workplace injury based events;

predicting, hours worked by said individuals with respect to a month and a year; and

extracting, from said current injury data and previously retrieved injury data, a demographics pattern with respect to occurrences of said event types.

5. The method of claim 4, wherein said identifying said event types comprises:

executing, a text analytics process and a hierarchal clustering process with respect to said current injury data to identify distinct injury type signatures;

evaluating a probabilistic map with respect to an injury signature of said injury type signatures that is associated, with said demographics pattern via a series of classification processes;

computing a severity or loss time impact of each injury signature of said distinct injury type signatures;

identifying a root cause with respect to each said injury signature via a text analytics process applied to an injury log;

reviewing information associated with said root cause; and

executing a set of time series models to predict an incident rate with respect to said injury type signatures.

6. The method of claim 4, wherein said extracting said demographics pattern comprises:

executing a random forest model with respect to extracting important demographic features of said individuals;

isolating a frequency of occurrence with respect to each event type based on a selected feature vector; and

executing a clustering algorithm to cluster said predicted figure workplace injury based events based on said frequency of occurrence across said event types.

7. The method claim 6, wherein said generating said injury risk mitigation actions comprises:

executing a integer linear program to extract a prioritized list of risk mitigation actions from a list comprising said distinct injury type signatures, wherein said executing said integer linear program is based on;

a predicted frequency, severity, and expected loss per event of each said injury signature;

an estimated cost of injury risk mitigation steps for each said injury signature; and

a process for maximizing a return on investment for each said injury signature.

8. The method of claim further comprising:

providing at least one support service for at least one of creating, integrating, hosting, maintaining, and deploying computer-readable code in the computing system, said code being executed by the computer processor to implement;

said monitoring, said storing, said generating said predicted future workplace injury based events, said generating said injury risk mitigating actions, and said generating said cost optimized reduction plan.

9. An embedded controller comprising internal dedicated electronic circuitry comprising a programmable logic array utilizing state information of specialized firmware program code configured to execute specialized for personalizing the electronic circuitry and a computer processor coupled to a computer-readable memory unit, and memory unit comprising instructions that when executed by the computer processor and the internal dedicated electronic circuitry implements a workplace risk factor identification and safety issue incident resolution method comprising:

monitoring, by said computer processor via execution of multiple geographically distributed dedicated hardware sensor devices, a plurality of workplace injury based events associated with a plurality of individuals at a multisite distributed workplace environment, wherein said dedicated hardware sensor devices comprise RFID sensors for identification, video camera based sensors for automatically recording said plurality of workplace injury based events, and mobile device type sensors for near real time reporting of said plurality of workplace injury based events;

tagging, by said computer processor, said plurality of workplace injury based events;

storing, by said computer processor within a centralized server, current injury data describing said plurality of workplace injury based events comprising tags from said tagging;

presenting, by said computer processor via a first clickable interactive view portion of a first GUI interface in response to activating said first clickable interactive view portion, a frequency of injury type at selected locations;

presenting, by said computer processor via a second clickable interactive view portion of a second GUI interface in response to activating said second clickable interactive view portion, injury locations associated with specified zip codes;

presenting, by said computer processor via a third clickable interactive view portion of a third GUI interface in response to activating said clickable interactive view portion, repair actions associated with said plurality of workplace injury based events;

performing, by said computer processor, a hierarchical clustering process with respect to incident keywords to obtain a set of distinct incident types defined by incident description text;

executing, by said computer processor, a set of time series models for each incident cluster type to predict a rate of future occurrence for each incident type of said set of distinct incident types;

executing, by said computer processor, a clustering algorithm to cluster demographic attributes by occurrence of frequencies across all identified incident types;

extracting, by said computer processor, demographic profiles of an active population of said plurality of individuals;

predicting, by said computer processor, a cluster id for an associated demographic profile of said demographic profiles;

normalizing, by said computer processor, each cluster level prediction by evaluating a proportional contribution of a demographic feature vector with respect to each said incident type;

combining, by said computer processor, said set of time series models to predict an incident type, a rate, and a detailed analytical model with respect to employee demographic features;

generating, by said computer processor based on said current injury data, said frequency of injury type at selected locations, said injury locations, said repair actions, and previously retrieved injury data describing previously retrieved historical workplace injury based events, predicted future workplace injury based events associated with future workplace injury based events with respect to a predicted plurality of individuals at said multisite distributed workplace environment;

generating, by said computer processor based on said predicted future workplace injury based events, injury risk mitigating actions associated with prevention of said predicted future workplace injury based events;

identifying, by said computer processor in response to said executing said clustering algorithm and via a text analytics process applied to an injury log, root causes for said plurality of workplace injury events with respect to injury signatures of previous work related injuries to identify distinct injure type signatures caused by activities monitored via said RFID sensor, said video camera based sensors, and said mobile type sensors;

evaluating, by said computer processor, a probabilistic map with respect to an injury signature of said distinct injury type signatures associated with a demographic pattern via a series of classification process;

identifying, by said computer processor, a root cause, of said root causes, with respect to each said injury signature via a text analytics process applied to said injury log;

executing said in injury risk mitigating actions such that said root causes are corrected thereby mitigating injury risks associated with workplace safety resulting in a maximal gain in productivity, incident risk resolution options, and reduced injury rates with respect to a frequency, severity, and expected loss associated with body stress events, motor vehicle accident related events, bodily injury related events, medical condition related events, biological factor related events, mental stress related events, chemical accident related events, and heat related events;

generating, by said computer processor based said injury risk mitigating actions, a cost optimized reduction plan for prioritized implementation of said injury risk mitigating actions; and

deploying, maintaining, and integrating computer-readable code said embedded controller for execution of said method associated with improving said incident risk resolution options and reducing workplace injury rates.

10. The embedded controller of claim 9, wherein said method further comprises:

transmitting, by said computer processor to a mobile device executing an interactive spatio-temporal application, said injury risk mitigating actions and said cost optimized reduction plan for presentation to a user.

11. The embedded controller of claim 9, wherein said predicted future workplace injury based events are generated with respect to future times of occurrence, future locations and frequencies of occurrence, and demographics with respect to said predicted plurality of individuals at said multisite distributed workplace environment.

12. The embedded controller of claim 9, wherein said generating said predicted future workplace injury based events comprises:

identifying event types associated with said predicted future workplace injury based events;

predicting hours worked by said individuals with respect to a month and a year; and

extracting, from said current injury data and previously retrieved injury data, a demographics pattern with respect to occurrences of said event types.

13. The embedded controller of claim 12, wherein said identifying said event types comprises:

executing a text analytics process and a hierarchal clustering process with respect to said current injury data to identify distinct injury type signatures;

evaluating a probabilistic map with respect to an injury signature of said injury type signatures that is associated with said demographics pattern via a series of classification processes;

computing a severity or loss time impact of each injury signature of said distinct injury type signatures;

identifying a root cause with respect to each said injury signature via a text analytics process applied to an injury log;

reviewing information associated with said root cause; and

executing a set of time series models to predict an incident rate with respect to said injury type signatures.

14. The embedded controller of claim 12, wherein said extracting said demographics pattern comprises:

executing a random forest model with respect to extracting important demographic features of said individuals;

isolating a frequency of occurrence with respect to each event type based on a selected feature vector; and

executing a clustering algorithm to cluster said predicted future workplace injury based events based on said frequency of occurrence across said event types.

15. The embedded controller of claim 14, wherein said generating said injury risk mitigation actions comprises:

executing a integer linear program to extract a prioritized list of risk mitigation actions from a list comprising said distinct injury type signatures, wherein said executing said integer linear program is based on;

a predicted frequency, severity, and expected loss per event of each said injury signature;

an estimated cost of injury risk mitigation steps for each said injury signature; and

a process for maximizing a return on investment for each said injury signature.

16. A computer program product, comprising a computer readable hardware storage device storing a computer readable program code, said computer readable program code comprising an algorithm that when executed by a computer processor of an embedded controller comprising internal dedicated electronic circuitry comprising a programmable logic array utilizing state information of specialized firmware program code configured to execute specialized software for personalizing for personalizing the electronic circuitry for implementing a workplace risk factor identification and safety issue incident resolution method, said method comprising:

monitoring, by said computer processor via execution of multiple geographically distributed dedicated hardware sensor devices, a plurality of workplace injury based events associated with a plurality of individuals at a multisite distributed workplace environment, wherein said dedicated hardware sensor devices comprise RFID sensors for identification, video camera based sensors for automatically recording said plurality of workplace injury based events, and mobile device type sensors for near real time reporting of said plurality of workplace injury based events;

tagging, by said computer processor, said plurality of workplace injury based events;

storing, by said computer processor within a centralized server, current injury data describing said plurality of workplace injury based events comprising tags from said tagging;

presenting, by said computer processor via a first clickable interactive view portion of a first GUI interface in response to activating said first clickable interactive view portion, a frequency of injury type at selected locations;

presenting, by said computer processor via a second clickable interactive view portion of a second GUI interface in response to activating said second clickable interactive view portion, injury locations associated with specified zip codes;

presenting, by said computer processor via a third clickable interactive view portion of a third GUI interface in response to activating said clickable interactive view portion, repair actions associated with said plurality of workplace injury based events;

performing, by said computer processor, a hierarchical clustering process with respect to incident keywords to obtain a set of distinct incident types defined by incident description text;

executing, by said computer processor, a set of time series models for each incident cluster type to predict a rate of future occurrence for each incident type of said set of distinct incident types;

executing, by said computer processor, a clustering algorithm to cluster demographic attributes by occurrence of frequencies across all identified incident types;

extracting, by said computer processor, demographic profiles of an active population of said plurality of individuals;

predicting, by said computer processor, a cluster id for an associated demographic profile of said demographic profiles;

normalizing, by said computer processor, each cluster level prediction by evaluating a proportional contribution of a demographic feature vector with respect to each said incident type;

combining, by said computer processor, said set of time series models to predict an incident type, a rate, and a detailed analytical model with respect to employee demographic features;

generating, by said computer processor based on said current injury data, said frequency of injury type at selected locations, said injury locations, said repair actions, and previously retrieved injury data describing previously retrieved historical workplace injury based events, predicted Ratite workplace injury based events associated with future workplace injury based events with respect to a predicted plurality of individuals at said multisite distributed workplace environment;

generating, by said computer processor based on said predicted future workplace injury based events, injury risk mitigating actions associated with prevention of said predicted future workplace injury based events;

identifying, by said computer processor in response to said executing said clustering algorithm and via a text analytics process applied to an injury log, root causes for said plurality of workplace injury events with respect to injury signatures of previous work related injuries to identify distinct injury type signatures caused by activities monitored via said RFID sensors, said video camera based sensors, and said mobile device type sensors;

evaluating, by said computer processor, a probabilistic map with respect to an injury signature of said distinct injury type signatures associated with a demographic pattern via a series of classification process;

identifying, by said computer processor, a root cause, of said root causes, with respect to each said injury signature via a text analytics process applied to said injury log;

executing said risk mitigating actions such that said root causes are corrected thereby mitigating injury risks associated with workplace safety resulting in a maximal gain in productivity, incident risk resolution option, and reduced injury rates with respect to a frequency, severity, and expected loss associated with body stress events, motor vehicle accident related events, bodily injury related events, medical condition related events, biological factor related events, mental stress related events, chemical accident related events, and heat related events;

generating, by said computer processor based said injury risk mitigating actions, a cost optimized reduction plan or prioritized implementation of said injury risk mitigating actions; and

deploying, maintaining, and integrating computer-readable code into said embedded controller for execution of said method associated with improving said incident risk resolution options and reducing workplace injury rates.

17. The computer program product of claim 16, wherein said method further comprises:

transmitting, by said computer processer to a mobile device executing an interactive spatio-temporal application, said injury risk mitigating actions and said cost optimized reduction plan for presentation to a user.