Sunday, September 12, 2010

The value of eco-system services for coffee production

For reasons that will become clearer in a few months, I've become somewhat interested in bio-diversity preservation and eco-system services. Examples of ecosystem services are pollination by bumblebees, decomposition of wastes, and flood mitigation and carbon sequestration by forests.

One of the main problems with ecosystem services is that it is hard to attach to them a monetary value. As a result, decisions to cut down forests, develop lands and interfere with water flows often undervalue these services. Even if there is a value attached to the service, the fact that it is provided by nature makes it harder to tell who should pay to preserve it.

I was recently shown a nice example of where the value of an ecosystem service has been quantified, an no less, in the area of coffee production! In [1], Ricketts et. al show the value of having a forest close to a coffee plantation. They conduct this study in a coffee farm in Costa Rica, and show that within a distance of 1km from the forest, the benefits of forest-based pollinators (i.e., diversity of bees) can increase the production of the farm by 20%. This observation, as they show, can be directly translated to a monetary value. The basic reason that the proximity of the forest is important is that the diversity of bees in the forest enable better cross pollination among plants (whereas, for example, honey bees typically focus on single branches when flowers are dense). Interestingly, the diversity of bees also reduced the number of peaberries produced, which may be slightly more controversial if your goal is to make money off peaberries (which some do).

Thanks to Gretchen Daily for sharing her article with me.

[1] Taylor H. Ricketts, Gretchen C. Daily, Paul R. Ehrlich, and Charles D. Michener. Economic value of tropical forest to coffee production. Proceedings of the National Academy of Sciences, August 24, 2004, Volume 101(34). Pages 12579-12582.

Friday, June 18, 2010

Fun with World Cup Soccer Statistics

As a teenager I was curious about which minutes in a soccer game are the most likely to have goals scored. I wrote a computer program that stored a database of all the goals scored in the Israeli soccer league for an entire year. I diligently went through all the sports sections of the newspapers and entered all the goals and minutes in which they were scored (feeling very mature that I was able to ignore my strong feelings about some of these goals). I calculated the statistics I was looking for, and the answer was: minute 65 was the most goal-rich minute.

Now, a few decades later, as the 2010 World Cup begins, I find myself asking the same question, or rather, revelling at how easy it is to capture the data, compute the statistics and share them with everyone in the world.

Using Google Fusion Tables, the tool developed by my team at Google, I created the visualization below. We're updating the underlying table as more goals are scored, so you'll always see the latest stats.

But that's not the end of it. Fusion Tables is a tool for data integration. We found some data on and joined it with our own table, and then created more interesting visualizations.

This one shows the height of the goal-scoring players. Read into it what you want.

This one shows the distribution of goal scoring among defenders, forwards and midfielders.

And finally, this visualization shows the clubs at which the goal scorers play.

Saturday, February 6, 2010

The Checklist Manifesto

I just finished reading "The Checklist Manifesto" by Atul Gawande, a very interesting book.

Gawande, a surgeon, essentially makes the following point. Given the incredible amount of knowledge we have accumulated in some professions, the complexity of certain tasks could be incredibly overwhelming to professionals (e.g., surgeons, airline pilots). Since in many situations these professionals work under pressure, they often forget some very simple yet important steps that later create unforseen problems (e.g., making sure the antibiotics are applied at a particular time before the incision is made into the patient).

Hence, he argues for simple checklists that teams should go through to ensure that important details are not glossed over. In the airline industry, checklists are used religiously. At every step of the flight, or whenever anything goes wrong, there is a checklist for the flight crew to follow. Gawande's main argument is that this principle should be applied in other professions as well, and in particular, in medicine. He describes his experiences launching such a checklist program with the World Health Organization and the impact that it had on reducing complications following surgery.

There are two main challenges this strategy. First, the checklist needs to be short as to not to completely slow down work. Hence, choosing and phrasing the items on the checklist requires significant thought. The second challenge is putting ego aside. For example, surgeons are used to being the kings of the operating room, and do not lightly take comments from nurses or other staff. Well, pilots have gotten over it, and they're not slackers in the ego department.

Gawande also gives examples from the construction industry and from restaurants, where constructing a high-rise or making sure that everything comes together at the right time on a customer's plate can be rather challenging. One main observation he makes from all of these examples is the importance of communication among the team members, in addition to the checklist. It is crucial for members of the team to communicate well with each other and building communication into the workflow is key. In that way, it's less likely that things fall between the cracks leading to additional problems.

To me the book was interesting because it points out that even if we build a huge body of knowledge in a particular domain, applying this knowledge in practice can be equally challenging.

A Trip to Australia

I recently returned from a trip to Australia, where I gave a keynote at the Australasian Computer Science Week, the annual gathering of computer scientists from Australia and New Zealand. You can see a journalist's account of what I talked about here.

There is a small but very strong database community in Australia, and I encourage anyone who has a chance to go down under and visit. The strength of the community was apparent when two of the three major annual awards were given for database work. Heng Tao Shen from the University of Queensland received the Chris Wallace Award. This is the top prize given for technical achievements across all fields of computer science (full professors are not eligible for this prize). Heng Tao made his mark spanning the fields of databases and multi-media.

The second award was the Ph.D Thesis Award that went to Michael Cahill who received his Ph.D from the University of Sydney under the guidance of my friend (and excellent cook!) Alan Fekete. Michael and Alan also received the Best Paper Award at SIGMOD 2008 for this work on serializable isolation for snapshot databases.

I was very fortunate to spend time with these winners. Heng Tao was my wonderful host in Brisbane and helped make a long-time dream come true -- sitting on a sunny beach in the middle of January (in Gold Coast). When I went to Sydney, Alan took me to an espresso machine making factory, where I got to see up close how these machines are made!

The coffee in Australia is amazing, and will be the subject of a different post. But if you're going to Australia and need coffee, check out my list of favorite cafes and you'll be happy.