Netbooks on the road

My part of this “netbooks” trial involved much hair loss. Since the base for my work with disconnected teenagers is a cybercafé, there is no obvious rôle for a small, pocketable computer in the normal context of what I do. To make good use of the opportunity, I had to let these machines go out of my control, into an environment where small high value objects are regarded as currency. The sponsors said they were willing to take the risk of loss, provided that I took what I considered reasonable care to minimise it … what, exactly, constitutes reasonable care when handing expensive stuff over to teenagers who may not come back, have class A drug habits, and are due in court on Wednesday for handling stolen goods?

The other question was what exactly to do with these machines, to justify taking the risk. These two issues were linked; my clients had to feel that something worthwhile was going on, if they were to respect the tools involved.

One subject which interests all of them, regardless of gender, is cars. A month before the netbooks arrived, I started discussing with them the relationships between weight, power, speed and acceleration in a car. They have rather more practical understanding of these matters than can be easily explained by legal experience at their age so I concentrated on trying to relate this to theoretical engineering models, first visual and then symbolic.

With the netbooks on hand, I brought the talk around to how we might investigate the actual (rather than maximum or advertised) speed and acceleration values for real cars in daily use. They were very interested in this idea, and were keen to try their hand at using spreadsheets for the purpose. Then they realised that they would have to write down a lot of information and bring it back to the centre, then key it in, before they could do anything with it; at that point, disappointment and loss of interest threatened. Like a good conjuror, I then produced the netbooks.

Gathering data

The scheme they devised involved teams of six, each team stationed downstream from a Pedestrian Light Controlled crossing (this allowed two teams per crossing, getting double data for each red light, at three different crossings). The team leader (let’s call her or him “A”) would stand by the lights themselves, and would have the computer with an open spreadsheet. “B” through to “F” would be at measured distances downstream from the lights.

When the lights turned red (probably because “A” had pressed the button, but I didn’t enquire too closely), “A” would take up a position beside the frontmost car and enter details (make, model including engine size if possible, number of occupants) into the spreadsheet. When the lights went amber, “A” would raise his or her arm and the others would prepare to start stopwatches (mostly on mobile phones, though a few used the function on their wristwatches). When the lights turned green “A” would drop the raised arm and start walking up the line; the rest of the team would start the stopwatches running.

As the lead car passed each team member, the stopwatch at that position would be stopped. As “A” reached each, the time on their stop watch would be entered into the spreadsheet. In this way, a database of timings at fixed distances for different vehicles was built up. The results were also visible in a predefined scatter plot at the right of the same screen, with an interpolated trend line, so the model could be seen developing as they worked. When complete, the sets of data were merged into a single sheet on the desk top and then filtered to compare different data for similar subsets.

As for the risk, I handed over the complete trial set to the two alpha primes in the group (one male, one female) and left them to arrange distribution; and all came back.

Taking it further

This probably seems an underutilisation of the equipment. The same data collection could, after all, have been done with a pocket PC or similar (in fact, the idea was partly suggested by Chandra’s Big Freeze which used Psion clamshells. But the experience of taking “proper computers” out, and being trusted to do so, was worth its weight in gold and stimulated desire to learn. There were, in any case, two follow ups which would not have been possible with handhelds.

First, there was use of a pure mathematics package to compare the experimental data with a theoretical model. Chandra and AbsentCat had described their use of SysQuake LE for projectile modelling. SysQuake is available for both Windows (in the cybercafé) and Linux (on the netbooks) so I installed both. Having set up a basic acceleration equation (d=½at²) on the PC, we set the value of a by trial and error to give a line which matched the spreadsheet data. The young people found this very empowering, and probably learnt more algebraic confidence in half an hour of SysQuake than in all of their time with me to date. They also learned, to their surprise, that most acceleration is over within a very short time (with speed surprisingly low and surprisingly constant) on urban roads.

Second, AbsentCat scrounged us the loan of a set of plug in USB interfaces allowing various types of switch to start or stop timers on the netbooks. The students had a lot of fun with trying out various switching devices. We were loaned some pressure mats which could be placed on the road, though too often the passing vehicles avoided them. We experimented with home made trembler switches, but they were too sensitive, and hard to position usefully. Lengths of rubber tube, filled with water, were laid across the road with light pressure sensitive microswitches plugged into the ends – these were the most successful, and supplied 95% of our usable data.

Broader benefits

The tremblers were a complete failure in data collection terms but worth their weight in gold for the interest which they provoked. A drop of mercury is placed in the bottom of a glass tube; one electrode is immersed in it, and another arranged as a circular collar around the inside of the tube, fractionally above the meniscus; any motion which shakes the tube causes the mercury to make contact between the two electrodes, completing a circuit. Most of my clients have, at some time, been involved in vehicle theft, and immediately realised the relevance of tremblers to car alarms. We got a lot of chemistry, physics and engineering time out of the resulting investigations – even starting a new set of data collection exercises to investigate the link between tube size, collar spacing, and the trade off between sensitivity and discrimination.

This second (more accurate) phase gave us enough data to further investigate the mathematical model, and to extend it into areas such as mechanical work or power/weight ratios. It also allowed us to compare vehicles by type (small car, four wheel drive, bus, lorry, motorcycle, etc). Most valuably, in some ways, it led on naturally to discussing the range of road behaviours exhibited by different users of the same vehicle.

[Contributed by BobTheBumbler]

Netbooks - initial hardware housekeeping issues

I have been using Psion and Palm pocket computers extensively for some years to place computer assistance in the hands of primary pupils “doing science” outside the classroom. Given a trial set of “Classmate” Asus EEE PC subnotebooks (or “netbooks”) for a month, my first concern was not their capability (obviously greater, and to be dealt with in another post) but how far they could replace their smaller equivalents in the same rôle. The two crucial issues, with small children, are portability and survivability.

Portability is a relative term. Many of the boys to whom I loan a palmtop machine simply put it in their trouser pocket. Girls, on the other hand, usually put it in a school bag along with their books and so on. These Asus machines are about twice the size of a Psion, four times that of a Palm device. That makes them unpocketable, but doesn’t much affect a school bag. For boys, then, a change in behaviour is often necessary for these machines to be considered “portable”, but not for most girls.

For that reason, I loaned all five machines out to boys on 24 hour tickets in the first week just to see what would happen. In most cases, they went into sports bags (and came back muddy) or satchels (and came back covered in grey fluff). In a significant minority (15%) of cases they were carried around continually in the hand, which places them at considerably greater risk (but see below on survivability).

After the first week they were loaned as required, regardless of gender or time span; as expected, the girls treated them exactly as if they were palmtops.

Survivability was more worrying, and I asked how much risk was acceptable in field trialling. The answer back from the sponsor was that deliberate attempts to test a machine to destruction would be unacceptable, but that we shouldn’t let potential hazard stop us from doing things we would do with a palmtop. It happened that a joint maths/sport project was under way, so the trial subnotebooks were added to the stock of Psions and Palms and allowed to go out onto football and netball pitches.

A football pitch provided the severest test of survivability. A pupil took one of the netbooks down to a practice game to try out both real time analysis of game descriptors entered into a spreadsheet (OpenOffice Calc, saving in Excel file format) and video capture to disk using the built in camera. The computer’s novelty attracted a lot of attention and it wasn’t long before attempts were made to take it away from its guardian, who resisted. In the resulting mêlée the computer was dropped, trampled on by several sets of studded boots and rolled over by half a dozen tussling nine year old boys. When order had been restored, the referee had to dig it out of the mud. Cleaning the mud out of USB ports, Ethernet socket, VGA output connector, sound jacks and, worst of all, the keyboard, took a lot of time, patience and cocktail sticks but, miraculously, everything was still in perfect working order. After that, we sealed all orifices with electrical insulating tape unless they were needed for use; proper sealing plugs would be better still, but would probably get lost fairly quickly.

Fast forward: despite horror stories like this, and my gut feeling that these machines are not ultimately as robust as handhelds, none came to grief in the time we had them.

My summary judgement: these are a valuable addition to the portable computing options available for primary science. Since Psion type machines are no longer made, and can only be replaced second hand, their gradual replacement on failure by these small subnotebooks seems a good strategy. At the same time, it would be a mistake to withdraw a working handheld (especially of the palm type). For as long as possible, keep existing palmtop hardware in use but expand enthusiastically with subnotebooks.

[Contributed by Chandra]

Experiments with a one-per-student computer

Asus’ EEE PC, though useful in many other areas (see more extensive review here), is a computer designed specifically for education. A wireless platform cheap enough, light enough, robust enough, small enough and powerful enough to be seriously proposed as a go anywhere, work anywhere, one per child point of wireless entry into a networked school system. We don’t know whether this vision is about to become reality at this moment, but we don’t doubt that it will come about in time – and the EEE PC is certainly closer than anything else we have seen to the keystone which would make it possible.

Over the past few months we have been sharing a set of these machines, moving them around different groups for a week or two at time and comparing notes on the results.

The machine is small enough to just about go into a handbag, as some of our young female teenage students demonstrated, is big enough for adapted touch typing after some practice, has on board wireless or wired network connectivity, is provided with three USB ports plus microphone/headphone jacks and is remarkable resilient.

Prices start at £167 (about $300 or €230 at time of writing), although the the ones we used were those with two or four gigabytes of storage at £220 or £250 respectively ($400/€300 or $450/€340). Each machine in our set was also provided with a one gigabyte SD/MMC card, on which the default documents folder was configured to reside.

Despite some remarkably rough treatment, the complete set survived and were returned to the supplier in full working order.

That’s it for now. We will follow up with individual posts on our separate experiences over the trial period.

[Contributed by Chandra on behalf of the whole trial group]