Thoughts on social media, business and journalism from Suw and Kevin Charman-Anderson

Enabling school students to do real science via CERN technology. Want to show kids that they can be involved in what is going on now, that science is vibrant and interesting subject.

The Langton Star Centre is ?at Simon Langton Grammar School, Kent, gives student change to work alongside scientists and engineers. They go to CERN every year, and one year visited the lab of Dr Michael Campbell’s Medipix lab, working on a chip for the ALICE detector. Can be used for medical imaging. Lots of research and collaboration using this chip.

Competition for schools to design experiment to go into space. Wanted to do a cosmic ray intensity detector using Medipix chip – called LUCID, which ?will fly in 2012. Won, but project was a bit expensive so they have an earthbound version too.

Wanted to get more schools involved, which led to CERN@School, so different schools can look at cosmic rays in space and on earth via detector in their own lab. Pick up data and then examine it in the school lab, can do particle recognition.

Found it hard to get the money together, but got a pilot scheme in a ten other schools that take data at a set time each day. Schools pool the data, then it goes up on LSC servers and can see it on a map. But how to analyse it and get good science out of it? Now have a model using grid storage and computing. Will soon be able to do analysis of tracks.

Next step would be linking up with other cosmic ray projects.

Expanded project would enable sophisticated analysis and potentially useful result. If had enough schools, would have an enormous network of detectors, might be able to discover particles above GZK limit.

CERN@School invigorates teachers as well as inspiring students. Hope to attract more scientists into schools. Doing real science, real analysis, is not only fantastic, it also shows how smart and capable these students are.

{ 0 comments }

Herbaria@home, herbarium records, snapshot of the world before agriculture, including areas now completely obliterated. Found plant, Ghost Orchid, thought to be extinct.

Plant don’t move, but they do invade, e.g. Oxford Ragwort. Scilian plant introduced in 1700 to an Oxford botanical garden, escaped, and now has spread out across UK. Roesbay willowherb, but railways have distributed seeds and now it’s everywhere. Plants also go by road, e.g. Danish scurvy grass, should be coastal, but now has colonised verges.

Plant populations are in flux. Modern survey data alone isn’t enough, so need the historic data to give context.

Web based project to catalogue old data. Collection of 50,000 documents he is working on, several million UK-wide, but even with willing volunteers (in person) there are too many records.

Online, Wikipedia established that people would do this sort of online work, you can allow open access editing and it wouldn’t be mayhem. Distributed Proofreaders showed that people will transcribe text from the internet.

Have taken photos of documents and put them online along with a form that people can fill out to say what they see in the label data around the specimen image, e.g. site names, collectors, date.

Some of the documents are quite clear, because they are printed, but there are a lot of hand-written documents, e.g. from 1859, and the handwriting poses quite a problem. Handwriting recognition may eventually get there but it is quite a long way off.

Once you have the data, can give it a grid reference and put it on a map. Can validate a lot of the data as they enter it. Need volunteers to collaborate and discuss what they see, so have active message boards. Have a pretty expert volunteer set, e.g. with plant recognition especially of rare plants.

Have worked on collections from several large universities and museums, and often they don’t have a full time curator for these collections so that data is inaccessible otherwise.

Peer-revuew of records, people have free access to edit anything, and public edit history for every record. Botanists able to spot errors and make changes, but a lot of non-professional botanical expertise, people keen to work on a project like this.

Some similar collections likely to come online in similar projects soon, e.g. insects.

Benefits: improve access to collections, raise profile of collections, and people enjoy it as a hobby.

{ 0 comments }

Einstein@Home is a traditional citizen science project. Have 100,000 computers at any one time contacting project and looking for work. People join the project by joining website, download & install software, and then leave it alone. Get a screensaver (which is very pretty!), and when their computer is idle it is analysing data.

Physics experiment data. Not simulating, but taking real data about physical world and searching for very weak signals that reveal neutron stars – very compact, small start, 10km radius, which beams radio waves like a lighthouse. As beam passes by Earth you see a flash. Forms when an ordinary star burns all its fuel and collapses under gravity, electrons get crushed into the nucleus, combine with protons to form neutrons, which are 100x smaller than the original atom. They spin very quickly for same reason an ice skater spins faster when they pull in their arms.

Example, Crab Pulsar, formed 1054AD, spins 33 times per second. About 100m neutron stars in galaxy, but have found about 1900, mostly near us.

Einstein@Home uses gravitational waves to search for neutron stars sent out by the star. Detectors, built in last 20 years, made of mirrors hanging from wires, and when a gravitational wave comes along the mirrors swing a bit. That can be detected.

Also use data from Arecibo radio telescope in Puerto Rico. First discovery: 11 July 2010. Signal was followed up the next day and reconfirmed quickly.

Found a second radio pulsar, currently unpublished, appears to be a binary system, but not yet clear what the masses of the stars are.

Publicity of first discovery has been very inspiring for users and project team. New users jumped when the publicity happened, and the number of users leaving the software running continues increase.

Square Kilometer Array, which will come online in 2019 or later will produce so much data that distributed computing may be the only way to process it.

{ 0 comments }

Most citizen computing projects can do no wrong. That does not apply to climate science. ClimatePrediction.net raises a few hackles, often in situations such as a recent meeting about investment in supercomputer, and when someone said, “I hear you can do a lot of this on PS3s nowadays?” there was a degree of hostility.

It’s just a way of addressing certain problems. People think of models as being done on a petaFLOP Cray XT-6, but most climate scientists don’t have access to these. They can have access to citizen scientists.

Climate modelling depends on:

• Complexity, e.g. number of processes, number of aspects of the sytem
• Resolution of your model, e.g. 100km scale, 10km scale or 1km scale
• Duration of the run
• Ensemble size, (groups of models) and this is where citizen cyberscience comes in

Often need to run models may times, and this is where ensembles come in.

Uncertainty in models varies. Uncertainty was felt to be underreported, so added subjective assessment of uncertainty, some of numbers are a little bit rounded, as they were decided on through discussion.

Suspected the model ranges were too small was because all the models matched the 20th century numbers ‘suspiciously well’. Need unrealistic models as well as realistic ones. You have to go outside the range that is fits perfectly in order to be sure you know what that range of forecasts are consistent with current observations.

Serious money to do a run, so they are looking for good models, not ‘bad’ ones.

By doing tests of different models (using citizen science), see that there -40 error bar was too pessimistic in terms of uncertainty, and the +60 was about right.

Learnt that the lower bound too low, upper bound about right, but this was through experimentation, not discussion. Therefore is testable.

What next? Using volunteer computing to see how extreme weather and climate are related, as global warming can cause both extreme hot and cold weather events, e.g. heatwave in Russian, Pakistani floods. Were they one event or two? Where they related to global warming?

Looking at the flooding in UK in 2003, simulating seasons where damaging weather events occur, both with and without the signature of climate change, to see if it had an effect. Looking for influence of external driver – human influence.

These are rare events, so have to model them many times to see if risk of extreme event has increased.

Projects in development, embedding regional models in simulations.

Have only used participants to provide compute power, so haven’t engaged participants brains. Big challenge faced is that only a few hundred people take part.

{ 0 comments }

Comp scientist at UC Berkley, building platform for citizen science. Looking for commonalities, software support that addresses community’s needs to make it easier for scientists to use volunteer power. Tech is only one piece of the solution.

Build platforms for:

• Volunteer computing
• Distributed thinking
• Education

Computational science:

Simulations are now so vastly complex that they can only be done on computer. Simulations at various scales,  e.g. proteins, ecosystems, Earth, galaxy, universe. . Need lots of computing power because need to fit models to observed data. To predict what’s going to happen you need to run thousands or millions of simulations.

Generation of new instruments, e.g. LHC, LIGO, SKA, gene sequencers, produce data at unprecedented rates, right at limit of computers to handle. Beyond limit of computers owned by unis or institutions. Science limited by computing power and storage capacity. What we need is not a faster computer, but higher throughput, i.e. a lot of computers.

Consumer digital appliances, e.g. computers, handhelds, set-top boxes, are all converging on similar hardware. Networks that connect them all: consuemr digital infrastructure. 1.5 billion PCs. Graphics processing improved through desire to watch HD TV and play realistic games, and GPU oft 100x CPU speed. Put there for games, but good for science.

Storage on consumer devices approaching the terabyte scale, network approaching 1 Gbps.

All this is ideal for science computing!

Compare consumer digital infrastructure with institutional counterpart, it’s way bigger and way cheaper than institutional computing. Supercomputers moving towards an ExaFLOPs in 5 years, but consumers already have 1000 ExaFLOPS today. Consumer spend $1 trillion per year!

BOINC, free open source software, anyone can create a project.

Utopian ideal, to have a lot of these projects, getting computing power by advertising research to the public, educating public on their project, so public supplies resources to science where they want to put it.

Boinc projects

• ~30 projects
• 300k vols
• 530k computers
• 3 PetaFLOPS

Volunteers can do more than run software. they can provide tech support, can optimise programme, translate website, recruit new users. Initially used message boards, realised that they weren’t working well for non-tech systems, so now have a system based on Skype, so people needing help can find someone who is willing to give that help via Skype.

Volunteers have a spectrum of confidence, and some users are malicious, e.g. they have had people trying to scam others users and trying to get their PayPal IDs.

Motivations study. People interested in doing science, want to show the world they have the fastest computer, people who want to be a part of a team.

Distributed thinking. Stardust@Home, interstellar dust photos, looking for grains of dust. People can do this better than computers. Interesting thing was needing to quantify accuracy of results. Created samples where they new the answers, i.e. either contained noise or had a particle. Every 5th image was a callibration images and so could keep track of false positives and negatives.

Also used replication, many people look at it, and then if there is consensus can look at calibration results and that shows if consensus is correct. Project found all the dust particles it could.

Created platform called Bossa. Middleware for distributed thinking, provides scheduling mechanisms, e.g. calibration jobs, replication. Open system with respect to assessment, scheduling policies.

Being used to find fossils.

Also extending Bossa to Bossa Nova, looking at more complex systems for asking people to do things involving creativity, problems for which there is no unique answer. E.g. complex problem solving, use volunteers to decompose problem into sub-problems, propose solutions, evaluate them, evaluate how a group of solutions might work together. Involves different skills. At software level, it uses people optimally, uses them for tasks to which they can contribute most.

Education and citizen science. If we can train people to do more complex staves we can achieve more. This is very important – if people learn more they may stick with a project longer, recruit more people to help, and you get more computing power.

Challenge of training or educating 100,000s people is a challenging problem, not attacked by traditional education theory. Heterogeneity is problematic: different backgrounds, education levels, locations, language. What makes this tractable is that there is a constant stream of students arriving, dozens, 100s, 1000s new users per day, so we have lots of people arriving interested in a course, so we can do experiments. If we haev two alternative ways of teaching a concept, we can rig up the software system to randomly show one lesson or the other and then they take the same assessment.

May learn one lesson is better than another, or one lesson is better for a subset, e.g. based on demographic or other attribute, and can then make an adaptive course where as we learn more about the student we refine how we teach them. Not just individual lessons, but overall structure of course.

Bolt: system for tailored education for large streams of volunteers.

{ 0 comments }

I’m here at the Citizen Cyberscience Summit for the next two days. Expect quite a few notes (although probably not all session!).

Lessons in crowdsourcing.

Nothing new under the sun. Science shaped by forces that have existed a long time, and we understand them. Works on calculation and making mathematical models work. That has long history as being citizen science, going back 200 years. Take large task, divide into small exchangeable jobs, send out into the world with instructions on how to do them. Charles Babbage wrote extensively about this in the 1830s.

Babbage was thinking about this because of his computing machine, how can you split big tasks up? Activity is largely about starting something, drive is towards radical organisation that moves towards convention, where you have people who lead, follow, varying levels of skill.

Babbage’s discussion, prime example he worked off, Gaspard DeProny, French Revolution, labour was very cheap. DeProny got 100 people to do maths tables, required for surveying.

~100 years later, 1875, post American Civil War, lots of people out of work, women who were widows, organised group of computers, i.e. women, who put together the Harvard Star Catalogue. Then again in 1907 by US Naval Obs.

And again in 1938 during the Great Depression – 450 people working at tables with paper and pencils doing calculations for scientists or government. Maths Tables Project.

A few people have adding machines, they are the leaders.

A NYC computing office, Columbia Uni Stats Computing Lab, 1930, had 6 employees.

Most of the Maths tables computers hadn’t been to high school, organised by arithmetic opersations e.g. – or +.

Often drew from poor classes at that time: blacks, women, Irish, Jews.

Planners had a doctorate or masters, operated 1938 – 1948, remnants existed til 1964.

Most scientists/engineers didn’t have access to computers until mid-60s when they had timeshare. Some not until the 70s. They worked with a worksheet, that was planned, a bit like programming. Early programmers were called planners, coding was called planning. Instructions, so workers didn’t need to understand what they were doing.

Built 28 volumes of tables, e.g. powers of integers, exponential functions.

Discovered there were particular skills, and started to look for specific calculations that were good for generating revenue, e.g. OSRD calcs, microwave radar tables, explosion calcs; LORAN navigation tables (precursor of GPS); general science calcs, e.g. Hans Bethe paper on Sun. First test of linear programming.

Labour economics. As you build skills, people want to use their skills and want to be rewarded for that skills. They want to advance. WPA studied labour and skill as a way of building identity.

Building skill

• Identity
• Accomplishment
• Avancement

Aspirational issues: Everyone wanted to be special.

Special Computing Group, had a room to themselves, had machines, almost all were women, and everyone wanted to move up to the special computing group. They started offering courses at lunch to develop those skills. Once they had completed that, were were opportunities outside. So lots of places they could apply those skills. Best measure of ability was the skill of the group.

They recognised that losing members of the group, breaking it up would damage the organisation. Started to have difficulty getting work, so soliciting work from scientists.

Gertrude Blanch, PhD, chief mathematician. Ran the computing group. Had to take everyone who came in the door, had to find ways to enforce discipline, e.g. people who didn’t think about carrying the ten, or who couldn’t concentration. Calculations were done 3 – 10 times each, didn’t just duplicate for sake of it.

“People doing hand calculations computing the same number the same way make the same mistakes” – Babbage

So did same calculation in different ways to ensure accuracy.

Crucial issue: How do citizen scientists relate to professionals? Professionals build walls around themselves.

National Academies of Science said they wanted to be of use to the gov’t, and help scientific work. WPA sponsored a lot of science. Internal comms of Nat Academies are ‘embarassing at best’. Group realised it was in a position it didn’t want to be in, and internal memos had one set of reasoning that repeats:

“Scientists are successful people. The poor, are not successful, because they are poor. Therefore can conclude that hte poor are not scientific. Ergo, maths tables project is not scientific, ergo their work is not good.”

NAS wanted the budget for the maths tables and wanted to do it “right” and ‘well”. But they could never have replicated it with students as budget wouldn’t cover it.

Handbook of Mathematical Functions: Largest selling science book in history.

Gertrude Blanch, finished PhD in 1934, was Jewish, was never going to be employed by anyone, but doing the maths tables project lead her to be employed by the Air Force, worked on supersonic air flow, jet nozzles.

The thing that we are doing is building skill amongst the general populous can never be overlooked.

Maths group had droped form 450 people to 120 as labour costs were higher. But claimed 120 was as efficient. by 1946, group had fallen to 60 people with specific skill, but still as efficient. People have titles, skills, identifiable expertise.

Lessons:

• We are creating skill, not just exploring the universe and doing science,
• Get people who want to be identified with project, part of their identity.
• Builds org with hierarchy, divisions of labour based on skill
• Encourage aspiration

DeProny users Adam Smith as justification, first 2 ch of Wealth of Nation, division of labour, identifying people with skill. Forces that shape these orgs and relationships that will support science, there is also a political economy that shapes it. which builds skill but divides jobs, creates leaders and followers. Must deal with science self-defining as enclosed domain.

{ 0 comments }