Saturday, 18 July 2015
23 September 2021
Interview with Dr. Mohsen Kalantari

“NASA International space Apps challenge in Australia: Melbourne University”

2012 May 12

Hossein Mokhtarzadeh

Edit: Katie Ewing

Reflection of My Mind

International Space Apps Challenge is a competition initiated by NASA in which space data is freely available to the public and scientists in order to solve challenges in different teams. This year, over 2000 people joined together and teamed up to solve one of the challenges. These challenges tackle problems that may affect every individual on the earth (link). On April 21-22, 2012, Melbourne was one of the 25 cities that participated in this competition. Melbourne University SDI center also took part this year and their app was selected as one of the best in Australia. About 50 apps were selected from all over the world.  I spoke with the leader of Melbourne University’s team, Dr. Mohsen Kalantari, about his experiences to find out how they could win the competition in Australia. See this video that shows how their app works.

Q: What was this competition about?

I received an email about this competition and the term NASA attracted me. I thought that this is going be about space and rocket science, but then I realized that we can contribute when I had a look at the previous challenges and winners. I saw that the winners were people with good ideas, so it’s not just about building something. In our research centre, we are doing things related to global issues like disaster management, such as floods and earthquakes, and I thought that we can contribute in this challenge.

Q: Although the idea was important, I think you would win if you could do it?

Yes, it is about ideas, but it is also about implementing ideas. NASA was trying to come up with solutions which have two characteristics.  First, it should be cheap, and then the implementation must be quick. We had to develop something from scratch. So we did not need lots of resources, only people with programming skills and ideas given a short period of time.

Q: How many people were in your team?

There were four of us. Myself, Amir Nasr, who is a research assistant in our center, Davod Shojaee, a PhD student, and Moein Ghodrati, a Master’s student from the computer department of Melbourne University.

Q: How about the other teams, were they all from Universities?

We were all from Melbourne University. A private company and also many individuals dropped in and formed a team without any previous plan. The private sector was prepared because they knew what they were going to do. And our team also knew what we wanted to do. But there were people who were interested in participating and as a result it was a good outcome for some of them, not in terms of implementation but because of good ideas that they could pursue.

Q: So what did you actually do?

We developed an app which harvests real-time data about disasters like floods or earthquakes or anything related to disaster from social media.  In particular, we implemented twitter and we visualized it in a user-friendly manner on a web mapping environment. If you tweet now that in Melbourne there is an earthquake, after about 30 seconds to 1 minute, it will come up on the map.  This is a collective intelligence of data, so you rely on people and visualize it. Either you are going to make a decision based on that or not. This is the authority’s decision or the people’s decision. But this is a knowledge-based one. We are mining data from twitter in real-time and this is the important part. There are applications that you can harvest data from twitter, but we are doing it in real-time.

If, for instance, any disaster happened now, in 30 seconds the whole world would know it.

Q: How do you do it real-time? Do you use Google maps for instance?

No, we developed an app which basically searches all current tweets and finds anything related to disaster. It locates that one and then gets the data, processes it, and creates a map of that. For example, if someone tweeted that there was an earthquake in Melbourne, we will find that one, get the data, and visualize it for the people. While there is a disaster happening here, in 30 seconds the rest of the world would know that there is something happening here.  If there is a need for aid, people can use this to provide help much faster.

Q: If there were a fake one, what would you do with it? How do you filter them?

The algorithms we have written try to filter the fake ones. For instance, if there is just one earthquake tweet, it means nothing. But if it gets 2, 3, 4 or 100 tweets, then we consider it seriously. There is a rough algorithm for that which needs fine tuning and further research. We just did it in two days for the challenge because we had to develop the idea. Behind the scene in terms of coding, it needs further work. But this could be a foundation that other researchers can work on it from.

Q: What is the next step?

The next step is that our project will be featured by NASA at the international level. There will be people from all over the world looking at these challenges and selecting them. Real judges and experts will select the best one.  There is also people’s choice.

Q: When do you think it will be announced?

I am not sure, perhaps in a couple of weeks we should know the results.

Q: If you win, will you be invited by NASA to showcase your challenge?

Yes, we will be invited by NASA to present the challenge in NASA.

«نوشته فوق می تواند نظر نویسنده باشد و الزامن نظر رادیو کوچه نیست»


TAGS: , , , , ,