Proceedings of the 21st International Conference on Distributed Computing and Networking, 2020
Disaster management involves intensive coordination among multiple agencies like police, fire dep... more Disaster management involves intensive coordination among multiple agencies like police, fire departments, public health, non-govt. agencies, including local volunteers/field workers. Accurate situational information about damage, resource needs, available resources etc., in the affected areas help the disaster management agencies in proper damage and need assessment and prepare suitable resource deployment plan. Crowdsourcing has become a popular approach for information collection where open crowds of people share multimodal situational information (text, images, audio, video etc.) about any event through social media posts. However, the authenticity and reliability of such posts are still debatable. Gathering situational data directly from the affected community (community-sourcing) can supplement social media posts to generate effective insights. In this paper, we attempt to design and develop a multiplatform disaster management information system where both social media-based crowdsourcing and community sourcing techniques are used to accumulate location-specific situational information. Subsequently, a coherent picture of the disaster situation is evolved through the integration of these local snapshots. Here, we explore how community participation, in the context of disaster management, can be enhanced through collaborative knowledge transaction, which eventually will lead towards the development of a resilient community. A field trial of our system is conducted involving a remote village community at Namkhana, West Bengal.
The opportunistic knowledge sharing scheme, proposed in Chap. 3, is used to accumulate the emerge... more The opportunistic knowledge sharing scheme, proposed in Chap. 3, is used to accumulate the emergency resource needs at the control station, over a smartphone-based DTN, for necessary resource planning. Due to typical characteristics of DTNs [1-3], the control station is likely to receive some corrupted or stale information. Even possibility of receiving no information cannot be ruled out. Thus, the opportunistically received resource needs are required to be thoroughly validated before these are used for resource allocation. Furthermore, resources in a post-disaster scenario are scarce and not all demands can be met. Minimizing the deficit in allocation of high-utility resources is crucial. On the other hand, since the underlying communication environment uses DTN, information about resource needs at the shelters reaches significantly late at the control station. Hence, minimizing the resource deployment time is inevitable. In this chapter, a resource planning mechanism is presented which is performed in two phases. First, a case-based reasoning (CBR)-driven need validation scheme that is executed at the control station is proposed. Next, a utility-based integer programming model is formulated using the opportunistically transmitted and CBR-validated resource needs for optimal resource allocation. The formulated model reduces the resource deficit as a whole and the total resource deployment time to achieve fast and effective disaster relief. The control-node (a computer/server at the control station) receives current resource demands from the forwarder-nodes, carries out a CBRdriven technique to validate/estimate the needs and performs a utility-driven optimal resource allocation strategy to minimize resource deficit and resource deployment time. Optimal resource allocation through CBR-driven need validation scheme and the integer programming model bring about efficient resource planning.
The most dominant challenges in post-disaster emergency resource planning are forecasting the exa... more The most dominant challenges in post-disaster emergency resource planning are forecasting the exact demand for emergency resources, communicating such demands to the control station, and validating these demands before using them for resource planning. Forecasting the exact demand for resources in a relief shelter becomes tricky, because situational parameters influencing these demands keep on changing. Moreover, the collection and transmission of demands of far-flung shelters are challenging, owing to the post-disaster disruption of communication infrastructure. All these lead to ad hoc allocation of emergency resources to the shelters. In this paper, we first derive a principal component regression model to forecast demand for emergency resources based on situational parameters at the shelters. Subsequently, we propose an opportunistic demand sharing scheme for gathering and disseminating resource demands to the control station using a smartphone-based delay-tolerant network (DTN). Finally, we suggest a case-based reasoning driven demand validation technique to ratify these demands and also to project the demands that do not get transmitted. Experimental results show that our system forecasts dynamically changing resource demands at the shelters with high precision, transmits near accurate demands to the control station, and perfectly validates the received resource demands.
Proceedings of the 17th International Conference on Distributed Computing and Networking, 2016
In a post disaster scenario, assessment of resource needs is indeed challenging because needs are... more In a post disaster scenario, assessment of resource needs is indeed challenging because needs are constantly changing and supplies trickle in at an uncertain rate. Moreover, information about resource needs of far-flung areas get rarely permeated owing to the disruption of cellular and other communication infrastructure, leading to a huge gap in perception about the actual needs in those areas. Therefore, a dynamic need assessment mechanism is essential to stay abreast of the actual situation. Movements of volunteers and relief workers are governed by two dominant characteristics of human mobility - spatial locality and spatial regularity. In this paper, we make use of such mobility characteristics to propose a knowledge sharing based approach for dynamic assessment of post disaster resource needs using a smart-phone based delay tolerant network (DTN). The proposed technique caters to the objective of bridging the gap in perception about actual resource needs.
Global Telecommunications Conference, . GLOBECOM . IEEE, 2003
Application of multipath routing techniques in mobile ad hoc networks has been explored earlier, ... more Application of multipath routing techniques in mobile ad hoc networks has been explored earlier, as multipath routing may help to reduce end-to-end delay, perform load balancing and consequently improve throughput. However, it has also been shown that the success of multipath routing in ad hoc wireless network depends on network topology and channel characteristics can severely limit the gain offered
The use of general descriptive names, registered names, trademarks, service marks, etc. in this p... more The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The demand forecast model and the utility function, derived in Chap. 2, are used to assess the ex... more The demand forecast model and the utility function, derived in Chap. 2, are used to assess the exact need for emergency resources at different shelters. As mentioned earlier, these assessed needs are required to be transmitted and accumulated at the control station for maintaining a fully functional relief chain.
Post-disaster need assessment deals with the accurate assessment of the need (i.e. demand and uti... more Post-disaster need assessment deals with the accurate assessment of the need (i.e. demand and utility) for emergency resource at the shelters. While demand signifies the amount of resource required, utility represents the exigency of that requirement. Due to lack of, or imprecise need assessments immediately after a disaster, relief requirements are generally set up based on coarse estimates by logisticians regarding what people would normally need. The effectiveness of this estimation depends on the competencies and experience of the logistician in control, often leading to impromptu allocation of typically scarce emergency resources. Thus, forecasting the exact demand and enumerating the correct utility of emergency resources are inevitable.
The CBR-driven need validation technique and the utility-based integer programming model for opti... more The CBR-driven need validation technique and the utility-based integer programming model for optimal resource allocation, elaborated in Chap. 4, bring about effective resource planning.
In this paper, we present a unified framework for both asset and environment visibility of an ent... more In this paper, we present a unified framework for both asset and environment visibility of an enterprise through real-time tracking and monitoring of enterprise objects using the Internet. The objects that are monitored can be personnel, equipment or an environmental condition. The architecture also enables the control of asset utilization by triggering actuators. Active RFID tags, combined with sensors and actuators, form a wireless mesh network based on IEEE 802.15.4. We develop a new routing algorithm oriented for these tracking applications where the network is sparse but comprises of extended legs. We elaborate a test-bed implementation for tracking underground miners and equipment along with monitoring the underground environment for possible emission of toxic gases.
The existing safety management systems are mostly manually triggered and reactive. Since real tim... more The existing safety management systems are mostly manually triggered and reactive. Since real time on-line data about the safety and security parameters are not available, there is minimal scope to take any prior preventive measure to avoid any kind of safety & security hazards. Since identification of any forthcoming crisis requires current information, and such information must be communicated in real-time, therefore the need arises for an integrated ubiquitous communication network for industrial safety, security and its management that provides efficient, reliable, secure exchange and processing of relevant information. The objective of the paper is to analyze the impact of using our Real-Time Tracking, Sensing and Management System using Active RFID, Sensors and Wireless Mesh Network for improving industrial safety & security. We will discuss a unified ubiquitous framework for both object visibility and environment visibility of an enterprise through real-time tracking and moni...
This paper presents a novel content-based query support for handling resources in fully distribut... more This paper presents a novel content-based query support for handling resources in fully distributed ad hoc wireless networks in the context of Disaster Management (DM). In a typical DM scenario, the most common problem is the inefficient management of resources (like food, drinking water, medicine etc.) due to poor coordination and lack of communication among the heterogeneous groups participating in relief operations. A proper approach thus, needs to be adopted for efficient management of relief operations. This should be capable of assimilating the information regarding inventory of resources from different participating groups and disseminate it over the distributed network in a fully decentralized manner. The major contribution of this paper is a two-layered approach of content-based searching using narrowcasting superimposed over a broadcasted information-fading layer. The objective of information-fading is to make a node aware of information content of other nodes in the network through broadcast-based multi-hop percolation of information. However the preciseness of knowledge about a node decreases or fades away with hop-distance, thus reducing the information overhead. The term fading of information actually implies propagation of progressively summarized information based on a semantic classification of information. In such a knowledgenetwork, our content-based query retrieval process becomes quite effective and uses narrow-casting (as opposed to broadcasting) to access the prospective destination through the relevant set of nodes only. Thus we have developed an intelligent multi-hop information retrieval system where the query itself will lead the search process to get the proper information from the nearest available destination. The per-hop narrowing of the search space will eventually reduce the unnecessary visits to irrelevant nodes. We present results from detailed simulation, showing remarkable reduction in the search domain by 2.5 times. We have also discussed the scalability of this model to larger networks.
Use of social media has established itself as one of the important information carriers in the fi... more Use of social media has established itself as one of the important information carriers in the field of disaster management. However, use of Twitter and Facebook by victims, first responders and others generates information that is varied, unstructured and unreliable. On the other hand, NGOs, operating in the disaster area, are often involved in intra-organizational communication using messaging apps like WhatsApp, and their group interactions can help in gathering meaningful data for situational analysis and need assessment. Our focus is to automate the process of filtering relevant information, query-based clustering of pertinent information from a WhatsApp group conversation of a specific volunteer group, so that situation analysis and need assessment can be done more rapidly. We have evaluated our scheme using WhatsApp chat log of a medical volunteer group in two post-disaster scenarios and concluded that it can provide valuable insights about region-specific resource requiremen...
The production of any mine is heavily dependent on safe interface between mining equipments and h... more The production of any mine is heavily dependent on safe interface between mining equipments and human beings. Protecting these resources and workers against possible accidents and maximizing their utilization through real time location monitoring and control can improve the mine safety and productivity to a great extent. At the same time, the sensitive areas of mining premises must be automatically protected against unauthorized intrusion to cope up with possible security hazards. Since identification of any forthcoming crisis requires current information and such information must be communicated in real-time, therefore the need arises for a wireless communication framework for mine safety, security and its management that provides efficient, reliable, secure exchange and processing of relevant information. The objective of our proposed work is to showcase the impact of using our Real-Time Tracking, Sensing and Management System using Active RFID, Sensors and Wireless Mesh Network f...
Proceedings of the 21st International Conference on Distributed Computing and Networking, 2020
Disaster management involves intensive coordination among multiple agencies like police, fire dep... more Disaster management involves intensive coordination among multiple agencies like police, fire departments, public health, non-govt. agencies, including local volunteers/field workers. Accurate situational information about damage, resource needs, available resources etc., in the affected areas help the disaster management agencies in proper damage and need assessment and prepare suitable resource deployment plan. Crowdsourcing has become a popular approach for information collection where open crowds of people share multimodal situational information (text, images, audio, video etc.) about any event through social media posts. However, the authenticity and reliability of such posts are still debatable. Gathering situational data directly from the affected community (community-sourcing) can supplement social media posts to generate effective insights. In this paper, we attempt to design and develop a multiplatform disaster management information system where both social media-based crowdsourcing and community sourcing techniques are used to accumulate location-specific situational information. Subsequently, a coherent picture of the disaster situation is evolved through the integration of these local snapshots. Here, we explore how community participation, in the context of disaster management, can be enhanced through collaborative knowledge transaction, which eventually will lead towards the development of a resilient community. A field trial of our system is conducted involving a remote village community at Namkhana, West Bengal.
The opportunistic knowledge sharing scheme, proposed in Chap. 3, is used to accumulate the emerge... more The opportunistic knowledge sharing scheme, proposed in Chap. 3, is used to accumulate the emergency resource needs at the control station, over a smartphone-based DTN, for necessary resource planning. Due to typical characteristics of DTNs [1-3], the control station is likely to receive some corrupted or stale information. Even possibility of receiving no information cannot be ruled out. Thus, the opportunistically received resource needs are required to be thoroughly validated before these are used for resource allocation. Furthermore, resources in a post-disaster scenario are scarce and not all demands can be met. Minimizing the deficit in allocation of high-utility resources is crucial. On the other hand, since the underlying communication environment uses DTN, information about resource needs at the shelters reaches significantly late at the control station. Hence, minimizing the resource deployment time is inevitable. In this chapter, a resource planning mechanism is presented which is performed in two phases. First, a case-based reasoning (CBR)-driven need validation scheme that is executed at the control station is proposed. Next, a utility-based integer programming model is formulated using the opportunistically transmitted and CBR-validated resource needs for optimal resource allocation. The formulated model reduces the resource deficit as a whole and the total resource deployment time to achieve fast and effective disaster relief. The control-node (a computer/server at the control station) receives current resource demands from the forwarder-nodes, carries out a CBRdriven technique to validate/estimate the needs and performs a utility-driven optimal resource allocation strategy to minimize resource deficit and resource deployment time. Optimal resource allocation through CBR-driven need validation scheme and the integer programming model bring about efficient resource planning.
The most dominant challenges in post-disaster emergency resource planning are forecasting the exa... more The most dominant challenges in post-disaster emergency resource planning are forecasting the exact demand for emergency resources, communicating such demands to the control station, and validating these demands before using them for resource planning. Forecasting the exact demand for resources in a relief shelter becomes tricky, because situational parameters influencing these demands keep on changing. Moreover, the collection and transmission of demands of far-flung shelters are challenging, owing to the post-disaster disruption of communication infrastructure. All these lead to ad hoc allocation of emergency resources to the shelters. In this paper, we first derive a principal component regression model to forecast demand for emergency resources based on situational parameters at the shelters. Subsequently, we propose an opportunistic demand sharing scheme for gathering and disseminating resource demands to the control station using a smartphone-based delay-tolerant network (DTN). Finally, we suggest a case-based reasoning driven demand validation technique to ratify these demands and also to project the demands that do not get transmitted. Experimental results show that our system forecasts dynamically changing resource demands at the shelters with high precision, transmits near accurate demands to the control station, and perfectly validates the received resource demands.
Proceedings of the 17th International Conference on Distributed Computing and Networking, 2016
In a post disaster scenario, assessment of resource needs is indeed challenging because needs are... more In a post disaster scenario, assessment of resource needs is indeed challenging because needs are constantly changing and supplies trickle in at an uncertain rate. Moreover, information about resource needs of far-flung areas get rarely permeated owing to the disruption of cellular and other communication infrastructure, leading to a huge gap in perception about the actual needs in those areas. Therefore, a dynamic need assessment mechanism is essential to stay abreast of the actual situation. Movements of volunteers and relief workers are governed by two dominant characteristics of human mobility - spatial locality and spatial regularity. In this paper, we make use of such mobility characteristics to propose a knowledge sharing based approach for dynamic assessment of post disaster resource needs using a smart-phone based delay tolerant network (DTN). The proposed technique caters to the objective of bridging the gap in perception about actual resource needs.
Global Telecommunications Conference, . GLOBECOM . IEEE, 2003
Application of multipath routing techniques in mobile ad hoc networks has been explored earlier, ... more Application of multipath routing techniques in mobile ad hoc networks has been explored earlier, as multipath routing may help to reduce end-to-end delay, perform load balancing and consequently improve throughput. However, it has also been shown that the success of multipath routing in ad hoc wireless network depends on network topology and channel characteristics can severely limit the gain offered
The use of general descriptive names, registered names, trademarks, service marks, etc. in this p... more The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The demand forecast model and the utility function, derived in Chap. 2, are used to assess the ex... more The demand forecast model and the utility function, derived in Chap. 2, are used to assess the exact need for emergency resources at different shelters. As mentioned earlier, these assessed needs are required to be transmitted and accumulated at the control station for maintaining a fully functional relief chain.
Post-disaster need assessment deals with the accurate assessment of the need (i.e. demand and uti... more Post-disaster need assessment deals with the accurate assessment of the need (i.e. demand and utility) for emergency resource at the shelters. While demand signifies the amount of resource required, utility represents the exigency of that requirement. Due to lack of, or imprecise need assessments immediately after a disaster, relief requirements are generally set up based on coarse estimates by logisticians regarding what people would normally need. The effectiveness of this estimation depends on the competencies and experience of the logistician in control, often leading to impromptu allocation of typically scarce emergency resources. Thus, forecasting the exact demand and enumerating the correct utility of emergency resources are inevitable.
The CBR-driven need validation technique and the utility-based integer programming model for opti... more The CBR-driven need validation technique and the utility-based integer programming model for optimal resource allocation, elaborated in Chap. 4, bring about effective resource planning.
In this paper, we present a unified framework for both asset and environment visibility of an ent... more In this paper, we present a unified framework for both asset and environment visibility of an enterprise through real-time tracking and monitoring of enterprise objects using the Internet. The objects that are monitored can be personnel, equipment or an environmental condition. The architecture also enables the control of asset utilization by triggering actuators. Active RFID tags, combined with sensors and actuators, form a wireless mesh network based on IEEE 802.15.4. We develop a new routing algorithm oriented for these tracking applications where the network is sparse but comprises of extended legs. We elaborate a test-bed implementation for tracking underground miners and equipment along with monitoring the underground environment for possible emission of toxic gases.
The existing safety management systems are mostly manually triggered and reactive. Since real tim... more The existing safety management systems are mostly manually triggered and reactive. Since real time on-line data about the safety and security parameters are not available, there is minimal scope to take any prior preventive measure to avoid any kind of safety & security hazards. Since identification of any forthcoming crisis requires current information, and such information must be communicated in real-time, therefore the need arises for an integrated ubiquitous communication network for industrial safety, security and its management that provides efficient, reliable, secure exchange and processing of relevant information. The objective of the paper is to analyze the impact of using our Real-Time Tracking, Sensing and Management System using Active RFID, Sensors and Wireless Mesh Network for improving industrial safety & security. We will discuss a unified ubiquitous framework for both object visibility and environment visibility of an enterprise through real-time tracking and moni...
This paper presents a novel content-based query support for handling resources in fully distribut... more This paper presents a novel content-based query support for handling resources in fully distributed ad hoc wireless networks in the context of Disaster Management (DM). In a typical DM scenario, the most common problem is the inefficient management of resources (like food, drinking water, medicine etc.) due to poor coordination and lack of communication among the heterogeneous groups participating in relief operations. A proper approach thus, needs to be adopted for efficient management of relief operations. This should be capable of assimilating the information regarding inventory of resources from different participating groups and disseminate it over the distributed network in a fully decentralized manner. The major contribution of this paper is a two-layered approach of content-based searching using narrowcasting superimposed over a broadcasted information-fading layer. The objective of information-fading is to make a node aware of information content of other nodes in the network through broadcast-based multi-hop percolation of information. However the preciseness of knowledge about a node decreases or fades away with hop-distance, thus reducing the information overhead. The term fading of information actually implies propagation of progressively summarized information based on a semantic classification of information. In such a knowledgenetwork, our content-based query retrieval process becomes quite effective and uses narrow-casting (as opposed to broadcasting) to access the prospective destination through the relevant set of nodes only. Thus we have developed an intelligent multi-hop information retrieval system where the query itself will lead the search process to get the proper information from the nearest available destination. The per-hop narrowing of the search space will eventually reduce the unnecessary visits to irrelevant nodes. We present results from detailed simulation, showing remarkable reduction in the search domain by 2.5 times. We have also discussed the scalability of this model to larger networks.
Use of social media has established itself as one of the important information carriers in the fi... more Use of social media has established itself as one of the important information carriers in the field of disaster management. However, use of Twitter and Facebook by victims, first responders and others generates information that is varied, unstructured and unreliable. On the other hand, NGOs, operating in the disaster area, are often involved in intra-organizational communication using messaging apps like WhatsApp, and their group interactions can help in gathering meaningful data for situational analysis and need assessment. Our focus is to automate the process of filtering relevant information, query-based clustering of pertinent information from a WhatsApp group conversation of a specific volunteer group, so that situation analysis and need assessment can be done more rapidly. We have evaluated our scheme using WhatsApp chat log of a medical volunteer group in two post-disaster scenarios and concluded that it can provide valuable insights about region-specific resource requiremen...
The production of any mine is heavily dependent on safe interface between mining equipments and h... more The production of any mine is heavily dependent on safe interface between mining equipments and human beings. Protecting these resources and workers against possible accidents and maximizing their utilization through real time location monitoring and control can improve the mine safety and productivity to a great extent. At the same time, the sensitive areas of mining premises must be automatically protected against unauthorized intrusion to cope up with possible security hazards. Since identification of any forthcoming crisis requires current information and such information must be communicated in real-time, therefore the need arises for a wireless communication framework for mine safety, security and its management that provides efficient, reliable, secure exchange and processing of relevant information. The objective of our proposed work is to showcase the impact of using our Real-Time Tracking, Sensing and Management System using Active RFID, Sensors and Wireless Mesh Network f...
Uploads
Papers by Siuli Roy