by Phavanhna Douangboupha
This is a brief and initial description of a proposal for resource pooling and scheduling systems during a disaster and/or crisis situation using a handheld device and the WWW technologies. The ideas are formed as a result of discussions with Prof. Jeff Sonstein (the Center for the Handheld Web, Department of Information Technology, Rochester Institute of Technology). The proposal is still at an early stage; hence all feedback and suggestions will be highly appreciated and are more than welcome. This proposal covers a lot of areas and will be narrowed down further after receiving feedback. If you are interested, please come back and read my next blog to see which direction (if I can decide) I will pursue.
The two main target user groups are emergency respond teams and general public. The following are possible scenarios of when the systems could be useful.
1. A medical doctor receives an alert that contains information about a disaster in the area on a handheld device. The doctor will then browse through the information provided on the device. The systems will provide information to the doctor as to the distance and time it would take to get to the location. The information will be updated in real time to alert users about any possible traffic issues and unexpected conditions, for instance, when roads are closed or cut off unexpectedly during the disaster. The systems will also display the current situation: an authorized user would be able to see the latest on-site resource information (e.g. how many doctors are already on-site). Upon the doctor’s arrival, the device will report his/her location directly to the central system, therefore eliminating the need for personal reporting. The information will be shared among disaster command centers. The tracking system will automatically be seen by authorized people such as the mayor or emergency personnel in charge. They would use the information to come up with an appropriate plan.
2. A crisis manager on- or off-site requires a resource management system to provide him/her with information needed to effectively assess the situation. Tiesyte, D., & Jensen, C. S. (2008, April) presents a technique for efficient cost-based tracking of scheduled vehicle journeys. They believe that geographical positioning and wireless communication technologies are useful to the managers and users of the fleets on vehicle status tracking. By the same token, data from existing centralized systems of resource pooling and scheduling systems will be highly critical during a crisis. It will support a specific subset of management decisions. For instance, how many medical doctors have already arrived on-site, how many more are required, how many more are on their way, and how much time it would take to get to the site?
3. A victim or a citizen who needs help, which could be as simple as needing evacuation from an affected area. During disasters, it is expected that wired communication, such as a landline, would be cut. Consequently, mobile communication is known to be a solution for such situations. Therefore, the technology should be easy to use; it should enable fast interaction between users and the systems during an emergency.
4. A citizen who responds to the crisis by sharing information and resources with other users. For instance, someone who happens to be in the same area and is willing to pick up neighbors who need to be evacuated. This would be particularly crucial in remote areas where help from emergency personnel may take more time to arrive. During high demand resource situations, sharing of resources among citizens (wherever possible and reasonable) will not only reduce the requirement for government/agency resources, but more importantly, it will assure the faster delivery of assistance. Having said that, there might be issues relating to trust and security which could be further investigated.
5. A father or a family member who uses the handheld device to locate or track their loved one’s current location. As suggested by Prof. Jeff Sonstein, during a disaster, it is possible for family members to get separated due to various reasons. The ability to track and communicate with members of their family could ease their anxieties somewhat.
6. Citizens who are interested in the current status of the crisis.
The idea of the project is to utilize handheld devices and WWW technologies combined with user generated content during a disaster or a crisis situation. The technology is a tool to support resource pooling and scheduling. To implement this system, it is required to have a web application to interact with users and accept or grab data from various sources. The data will then be analyzed to feed into the central data source. Consequently, suggestions (predictions) and plans for the resource sharing and pooling will be generated. For example, the system shall have an algorithm to predict, based on data, a near-future resource requirement such as how many medical doctors will be needed or how many rescue vehicles will be required. The number of vehicles to send to suburb A might be less compared to suburb B, if people in suburb A have already been evacuated, or already have enough shared resources (cars) among them. Additionally, information about which cities can pool resources and the estimated travel times to deliver the resources will help a mayor or a chief of police address the situation more effectively. The project will make use of existing technologies such as GPS systems, Google map, Ushahidi resource, and the location-aware routing prediction method such as in Zhang, Q., Mayes, K., & Markantonakis, K. (2005) and Tiesyte, D., & Jensen, C. S. (2008). The system will provide information regarding possible traveling routes from point A to point B on a handheld device and give notification on any problematic traveling routes for users, especially when urgency and timing are highly critical. The sensitive information will only be shared among authorized users such as policemen, medical personnel, firemen, rescuers, emergency teams, etc. This information will be pulled from all existing databases. Examples of such information include personnel information and city resource information. Other kinds of data required include scraping from other online existing sources that will not be stored on a database. Nevertheless, a derived report or the result analysis data could be stored on the database for further situational analyses. Lastly, the handheld device should have a simple user interface not only to provide information to users but also to allow users to share their information. These are the three possible sources of data – private or government data, data that are already existing online (e.g. Ushahidi), and user-generated contents.
Possible Users:
- A medical doctor, a policeman, a fireman, emergency personnel, etc.
- A mayor or a police chief on location or off-site
- A citizen in the area who needs/requests help
- A citizen in the area who is available to help/has resources to share
- A citizen who is interested in the situation or needs information about crisis for various reasons.
Some Advantages:
- Accessibility – mobile
- Sharing data and information anytime and anywhere; data sharing among citizens – user-generated content. For instance, an accident might block traffic on a small road where it is less likely to be reported. This real-time information could be shared among users.
- Scheduling and sharing resources
- GPS location
- Prediction systems to identify the best resource sharing and pooling to report to a decision maker
- Prediction systems to provide several directions to users according to their GPS’s location
Possible Connection Structures:
- Client to Client
- Client to Server
- Server to Client
- Server to Server.
Some Issues/Concerns:
- How to integrate data from different sources?
- How to check/confirm data integrity from user-generated content?
- What algorithm to use for the prediction systems?
- System Security Issues
- Is the scope of the project too broad/big? Basically, this idea can be seen from the perspective of resource management. But it also includes the general user’s functionalities during a crisis.
Some Current Technologies:
- An intelligent and situation-aware pervasive system to support debris-flow disaster prediction and alerting in Taiwan2.
- A wireless First Responder Handheld Device for Rapid Triage, Patient Assessment and Documentation during Mass Casualty Incidents1.
- Cost-based tracking of scheduled vehicle journeys3.
- Ushahidi [An opensource project for handheld device where people can submit crisis information to the website]4.
References (Annotated Bibliography)
1. Killeen, J. P., Chan, T. C., Buono, C., Griswold, W. G., & Lenert, L. A. (2006).
A Wireless First Responder Handheld Device for Rapid Triage, Patient
Assessment and Documentation during Mass Casualty Incidents. AMIA Annu Symp
Proc, 429-433. Retrieved September/October 30, 2008, from American
Medical Informatics Association Web site: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1839472
This article describes systems that provide wireless handheld device solutions with an electronic medical record in a mass casualty incident or a disaster. This paper shows how important accessible and timely information is during a critical situation where essential communications are highly important. Consequently, the technology is shown to be a valuable tool in particular where there are limited resources.
2. Kung, H.-Y., Ku, H.-H., Wu, C.-I., & Lin, C.-Y. (2008, January). Intelligent and situation-aware pervasive system to support debris-flow disaster prediction and alerting in Taiwan. Journal of Network and Computer Applications, 31(1), 1-18 . Retrieved September 12, 2008, from The ACM Portal 2008 Web site: http://portal.acm.org/citation.cfm?id=1321783.1321931&coll=&dl=ACM
This is one of the most current academic papers in the field of disaster prediction and alerting mobile Internet communication. The article contains figures of the network and system architecture that provide a broad perspective on how the systems can be implemented using mobile appliances and the Internet. Consequently, this shows a solution of how mobile appliances and Internet technologies can be applied to solve a real life problem.
3. Tiesyte, D., & Jensen, C. S. (2008, April). Efficient Cost-Based Tracking of Scheduled Vehicle Journeys. IEEExplore, 9-16. Retrieved September 12, 2008, from The Ninth International Conference on Mobile Data Management Web site: http://ieeexplore.ieee.org/iel5/4511414/4511415/04511429.pdf?…[more]
This paper provides an example of how a mobile device solution can be used in developed countries. This provides a different point of view on the topic. It is also a technical paper that contains a general algorithm for the vehicle journeys system and its experimental results. This paper also states several future research directions on the topic.
4. Ushahidi [An opensource project for handheld device where people can submit crisis information to the website.]. (n.d.). Retrieved September 11, 2008, from Ushahidi.com Web site: http://www.ushahidi.com/
This site has links to articles and current solutions for mobile disaster warning systems. The website is also a forum where the technology discussions can be viewed.
5. Zhang, Q., Mayes, K., & Markantonakis, K. (2005, November). A user-centric m-payment solution. IEEE Mobility Conference 2005. Retrieved September 10, 2008, from IEEE Xplore Web site: http://ieeexplore.ieee.org/iel5/11013/34698/01656674.pdf?…[more]
The article proposes a user-centric design approach for mobile payment technology. From the solution, several technologies can be identified. It provides a technical solution which can be used to investigate how to develop the technology.
Links to other blogs in this project
- Resource pooling and scheduling systems during a disaster and/or crisis situation using a handheld device and the World
Wide Web (WWW) technologies - Web-based resource tracking system implemented on a mobile device
- Location Tracking Techniques
- Creating REST Web Service
- REST Client simple Implementation and Test
- REST Service Implementation using HTTP – Cross Domain Request
Tags: location tracking, mobile device
Sounds interesting. You may want to look at what the TSF people are doing too, as they are first on the ground for this sort of thing. http://www.tsfi.org/tsfispip/spip.php?lang=en
Good luck.
Bruce makes a very good point. TSF has a good deal of experience setting up telecom infrastructure, to service NGOs and international organizations as they respond to emergencies.
jeffs
Nice write-up. I think that it’s good, but I’m worried that it might be trying to reach too far at first. Just from what we’re building at Ushahid, I can see that trying to build that type of software with all the functionality that you are suggesting would take quite a bit of time. What might work though, and would have everything you discussed, would be a modular system of working solutions that can be put together like Lego blocks for different types of scenarios.
Take for instance the work that InSTEDD is doing around GeoChat and Mesh4x, and then see how that can work with citizen generated information using a tool like Ushahidi. Add to that data from other public (WikipediaMaps/Twitter/FaceBook) and government sources and you have a pretty good idea what’s going on.
There are always a couple challenging issues though. First, the aggregation of data – which is why we need to come up with some basic agreed protocols. Second, the visibility of data to disaster responders versus the public – what’s the granularity and who decides?
A third point to consider, or maybe I just need clarification on it is whether this is for the US only or for other nations around the world. We can believe that the US government is generally benign in a crisis, that’s not the same the world over. So, a centrally (gov’t) controlled system like this would not always be favorable in other parts of the world.
Now, at Ushahidi we’re trying to build the core platform to be usable through extensions with other software. We think that what you’ve just described is possible, but that there might need to be a way to distance yourself from the use of the software. Basically, provide basic tools and instructions that can do what you suggest – a toolbox if you will – then get out of the way and let the locals manage it.
ehersman makes some good points. I particularly appreciate (1) the idea of taking a modular approach, and (2) being certain people understand this is not a US-only effort.
jeffs
I pretty much agree with Erik’s comments here (and he does have a lot of experience building things like this). My thoughts are as follows:
1. It seems like you’re looking to build a carefully defined solution to a problem, rather than a platform which could be used for the purposes you define. In my experience, it may be better thinking more in terms of building a platform which has the flexibility to let users ‘do the kinds of things’ that you think they may want to do. If it’s flexible enough then they’ll end up doing things that you never expected, which is where it gets really exciting from a technical/usability standpoint
2. You have many outputs listed here, and only a few listed sources for data inputs. Do these exist, and how accessible are they? Are the systems you’re looking to rely on for your inputs as resilient than the one you’re looking to build? It would be no good if your service is all nicely up and running and the inputs have all crashed or gone down
3. Your assumption that “During disasters, it is expected that wired communication, such as a landline, would be cut. Consequently, mobile communication is known to be a solution for such situations” isn’t always true, or going to be true. During the London terrorist bombings, for example, the mobile/wireless networks ground to a halt with the traffic (people texting loved ones, trying to call, etc)
4. Are there specific problems with crisis reporting/administration that you’re trying to solve? What doesn’t work now, and why doesn’t it work? Are you clear about what value this service may add, and/or is there a danger that you may be re-inventing wheels in some cases?
This sounds like a great and interesting project. I’ll look forward to seeing how the idea evolves which, to me, will be almost as interesting as the end product itself.
Best of luck with it!
Ken
http://www.kiwanja.net
Very useful and focused comments, Ken. Thanks.
The comment about designing and building a flexible platform seems to me particularly correct. Each emergency situation is going to be unique. Creating or extending a platform to offer capabilities rather than going for pre-set solutions may be a very good design approach.
I am also of the opinion that starting on the input side may make some sense here. What kind of inputs should it handle? how resilient are they? who is likely to provide them?
I very much appreciate you giving her your time, and think Phavanhna will end up with a good and useful project this way.
jeffs
Some thoughts… First, while I’d really really like to subscribe here with my feed reader, I’m not actually interested in reading essays per se. Blogs have their own format/wording. (Nuke the references, make them little hyperlinks. We don’t need more than that.) Move the phrasing more informal.
Onto diaster commo. The sine qua non for disaster communications is graceful degredation. First, most of the network will be dead as a doornail during a real disaster. During a not-really-real disaster (i.e. minor disaster blown out of proportion via media) it’ll be dead due to network segregation. Second, there are a number of gubermint infrastructure tools that will activate automatically. While these are nice (make sure to talk to local EMS people about the targeted evacuation calling. Roch has a decent system.) they will also play merry hell with commo.
So, the first thing to do is step back for a moment and ask yourself(s) what kind of disaster you’re talking about. That will then inform what resources are available. (Remember, duration is important as mobile phones don’t last forever, a fairly significant number don’t recharge frequently, and computers run of mains power).
Also, my advisor Deborah Bunker has actually looked at these sorts of questions for Sydney. Try to look up some of her papers.
Once you have that, then build to take advantage of what resources you have.
Thanks, Brian. I think you are absolutely correct about the style issue. A less formal style will attract more readers and thus more feedback for Phavanhna.
I believe Phavanhna has started to try to talk with the Telecom Sans Frontieres folks (http://www.tsfi.org/) about how things work in real disasters, and I hope to connect her with some more governmental folks too.
Your suggestion re your advisor’s work is great, and I expect Phavanhna will follow that one up. Thanks a lot for your useful feedback to her.
jeffs
Dear Bruce,
Thank you for the link. I am looking at what they are doing now.
I hope to come up with a more detail description on what i want to do. One of which i will break down my idea into two to three possible sub which i will choose to take one.
Dear Ehersman,
Thanks so much for your thoughtful suggestions.
I talked to Prof. Sonstein today and we really like your suggestion of modular approach.
My idea is not only target the US. I am from Laos and in my country the number of mobile phone distribution is higher than computer. Also, we have many remote communities where i believe this would benefit them. For instance, this could help the small communities to reach out to each other (near by city). Other example would be like the case of TSF like Bruce suggested.
I like what Ushahidi is doing and I have a high hope that i can learn from what you guys are doing. More importantly, i hope that collaboration could be formed. I know that Prof. Sonstein is helping me to make the connection.
Dear Ken,
I am so excited that I receive many good feedbacks and suggestions from you guys. Thank you, Ken.
My next assignment is to take all the advices and try to break my idea down to two to three main directions. I then will have to come up with a more detail framework on each direction and I will start by trying to find answers to the questions posted here or at least a long that path.
I will have to give more thoughts in to this.
Again, thank you.
Phavanhna
Thanks, Brian.
I tried to write something between a blog and an essay. It was a good way for me to practice my writing and I believe it would help me to start my real thesis proposal which would have to be an essay type. I talked to Prof. Sonstein today and he gave me a similar suggestion. So, for my next blog I will try to write a real blog.
Thanks for the suggestion about the infrastructure and the hardware issue itself.
And, I will definitely look up for Deborah Bunker papers.
Phavanhna
I got a note from Katrin Verclas, Co-Founder & Editor of MobileActive.org:
“This exists — Sahana? http://www.sahana.lk/”
It is worth looking at, although the fact that there has been no update to the package for the past 10 months does worry me a bit.
jeffs
Dear Prof. Sonstein,
I just quickly checked Sahana’s download links. Most of them seem to work. I will investigate more about it later.