The great experiment

by Clive Shepherd


Training network specialists is an expensive business, whether you use a classroom (expensive to equip), simulations (expensive to build) or on-job learning (potential expense of errors off-the-scale). Until now. Remote, online labs allow learners to experiment to their hearts’ content with real hardware and software, solving real networking problems. Someone else buys the kit, someone else worries about what happens when it’s broken. Too good to be true? In this article, Clive Shepherd talks to the men in white coats to see just how close e-labs get to replicating the real-world experience, without the headaches.

Contents
Learning by doing
So what is an e-lab?
Options for network training
Blending e-labs into the curriculum
E-labs in action
Resources

Learning by doing

After more than a hundred years of research into adult learning by some of the world’s most eminent psychologists, we have discovered a few things that we can apply with confidence to the real world of education and training. Like the fact that adult learners want to be in control of the learning process; that they prefer content that is relevant to their current working lives; that they benefit from learning that is enjoyable; and that they like to learn through experience wherever possible. You might say this is just common sense, but evidently not that common as a great deal of the training that we dispense continues to be abstract and irrelevant.

The authenticity of a learning experience is a critical element because it affects the transferability of learning, at least according to the constructivist approach to learning. Simply put, it’s more difficult for a learner to apply an abstract concept to their work than a skill they have practised in a life-like or real-world setting. It’s possible to look at authenticity on a scale. At the ‘very low’ end of the scale we have abstract theory, that you’d find in a book or a lecture. Moving up to ‘low’ we find depictions of real-world situations, such as demonstrations, modelling, worked examples and case histories. At the ‘mid’ point are activities based on real-world situations or opportunities to share real-world experience, through case studies and discussions. On to ‘high’, where we find life-like but artificial activities, such as role-play exercises in a classroom or a simulation delivered from CD-ROM. At the ‘very high’ end of the scale is learning on-the-job learning; for example action learning or on-job instruction.

Readers of TACTIX are not stupid and by now you will have realised just how hypocritical this article has become. You wanted to read about e-labs and I’m bogging you down in so-called learning theory. A fair point, so let’s see how this applies to the problems of the training of network specialists. Well, no-one’s saying that technical training won’t benefit from activities that are lower down the scale – a little background theory, a demo or two, an opportunity for discussion with your peers and maybe even a simulation; it’s just that these don’t go quite far enough. At some point every learner needs to get their hands dirty and, as any small child will tell you, that can be a lot of fun. According to Roger Schank, author of Virtual Learning: ‘Listening to endless lectures and memorising countless facts and figures aren’t fun activities. What’s fun is doing.’

So what is an e-lab?

Before looking at how e-labs might fit into a technical training curriculum, let’s first clarify just what they are. ‘E-labs’ are ‘remote labs’, which are also ‘live labs’ – unfortunately there is no accepted, generic name as each vendor applies their own branding. The underlying principles are, however, broadly the same. E-labs allow learners the ability to interact freely with real network hardware and software over the Internet. Typically the learner is provided with a written description of a task to carry out and is then able to either configure network hardware or interact with server software in order to carry out the task in much the same way that they would if the equipment was sitting right in front of them.

Graham Shenton is Managing Director, EMEA, for e-labs pioneer, Logilent: ‘You can think of it as a giant timesharing service. We have a huge room full of kit including all the hardware and software that any student could need. For the Cisco curriculum, that includes some 600 routers, switches and so on. Students simply request the hardware that they need and this is allocated to them for the duration of their session, typically about 90 minutes.’

Logilent’s Founder and VP Development is David Clarke. He explains how the service originated: ‘We decided to use real hardware and software in our e-learning offerings back in 1994, since when 250,000 lab sessions have been completed. Hardware has been the most challenging problem, particularly as user volumes grow, but we now use a dynamic switching matrix so students can develop their own equipment configuration on the fly. This improves scalability by 400%, while providing students with the same feedback they would obtain in a live environment.’

Ray Geoghegan is VP e-Learning, EMEA, for Global Knowledge, whose ‘Remote Labs’ service was launched a year ago: ‘We have banks of servers operating under Linux, NT, Windows 2000 and Unix, with XP, .NET, SQL and Exchange Server coming by the end of Q3. We have developed our own partitioning software to allow many different students access to the same computer.’

E-labs use online application sharing technology to make it possible for learners to interact with remote servers. This may start the bandwidth alarm bells ringing, but Shenton explains: ‘The labs work fine at 56K, but the interaction with the software is more realistic at double this rate or better. However, the Cisco labs work absolutely fine at 56K.’

Providing students with the ability to experiment as they wish with someone else’s hardware and software sounds risky. What happens when they bring the system down? Shenton: ‘It’s true that students can do as much damage as they can in a live environment, although they tend in practice to be quite responsible. Nevertheless, we have put in as many failsafes as we can, so we can usually recover from a crash in 10 minutes.’
So, e-labs would appear to be accessible, scalable and reliable, which doesn’t sound like any networking environment I know (only kidding). But how do they compare with the training solutions we’ve all grown to love?

Options for network training

Currently, as a provider of training solutions, you have a number of options for providing students with practice in working with network hardware and software. The simplest option is to allow them to practise with your organisation’s current network, trying out various options and learning from their mistakes; but then the simplest options aren’t always the best, as you’d soon realise on your way to the dole office after the inevitable major systems crash.

Your second option is to set up all the hardware and software you could need in a classroom, where any damage that is done is confined within the walls of the training centre. For some time this has been the most popular option and one which all the major IT training providers have been happy to offer. However, keeping up with all the latest hardware and software can be a problem, as Clarke explains: ‘Networking equipment is getting more and more expensive. To provide a full Cisco curriculum would require $1.5m of equipment and hours of setup time for each lab. And if you need to move the training from one location to another, you’re talking about a major shipping challenge.’

A third option is to simulate the hardware and software in some way, as part of a packaged CD-ROM or web-based training solution. But simulations aren’t quite the real thing, as Geoghegan has realised: ‘Feedback from students tells us that simulations, whilst useful, have their limitations. Because they never contain the full functionality of the live system, it is actually quite hard for students to make mistakes. And without the opportunity to make mistakes, your learning is limited.’

So, this is where e-labs come in, providing you with a fourth option. Clarke: ‘The only thing that an online lab cannot replicate is physically plugging in the equipment. Our own research shows that 90% of the real learning takes place with the console/GUI and only 10% is physical. However, we recognise that this 10% is important and so we offer it through third party classroom providers. The advantage for them is that they only need a minimum of hardware to cover this need.’ And the effectiveness of e-labs has been borne out in practice, as Clarke explains: ‘An industry average pass rate for vendor certifications is 55%, less if the student is learning from a book. E-learning can raise the pass rate to 75%, but with our integrated solution, including extensive use of online labs, we’re achieving 94%.’ And, according to Geoghegan, e-labs actually provide a more realistic experience than in the classroom: ‘The feedback you receive can actually be more aligned to the real world than it would be live. These days engineers will frequently have to diagnose problems online before making a visit in person.’

Blending e-labs into the curriculum

One thing is for certain and that is that e-labs do not represent a stand-alone option – like so many new training tools, they need to form part of a blended solution. So how are vendors integrating e-labs into their curricula? Sharon McKechnie is an e-Learning Consultant for Spring, who has been using e-labs since March: ‘We like e-labs because they are real-world, unlike simulations. In conjunction with our partner, Productivity Point, we offer e-labs as an optional ingredient alongside classrooms and generic e-learning from NETg. We’ve found that the people who respond the best to the concept are the techies themselves. They really like the ability to get hands-on practice, which you can’t obtain with self-paced e-learning alone. These are people who like to roll their sleeves up and e-labs allow them to do that any time of day and night, from wherever they are.’

At Global Knowledge, e-labs are available to students regardless of their preferred learning medium, as Geoghegan explains: ‘We develop our classroom, self-paced e-learning and virtual classroom courses side-by-side, so they all share the same content. We’re already using e-labs to support our online offerings and we expect to extend this to the classroom as well, to reduce the amount of distributed hardware needed at each location.’

Logilent have been using e-labs for sometime, as part of a predominantly e-learning package, but have more recently opened up the e-labs as a separate offering for use by individual learners, training departments or resellers. Clarke: ‘As an example, unlimited access to all 112 Cisco labs for a year would cost an individual about $1000, whereas a single lab for a single person would cost $50. By contrast, Cisco themselves purchase from us and then offer labs free to all their channel partners.’

Other providers have constructed new blended solutions, but chosen for now to leave e-labs out of the equation. Brian Sutton is Chief Educator for QA Training: ‘In designing our new blended MSCE, we evaluated a wide range of alternative methods. We chose to go with a combination of web-based training from Microsoft, which incorporates guided practice using simulations; virtual classrooms for tutorials; and real face-to-face classrooms, exclusively for problem-solving activities in teams, based on real workplace scenarios. We felt we could achieve what we wanted better using this approach, but are keeping an eye on e-labs as a possible option for future courses.’

E-labs in action

So what impact can we expect e-labs to make in the UK market in the foreseeable future? Well, there’s no doubt that we’ll be hearing a lot more about them, as an increasing number of UK providers bring solutions to market, more often than not in conjunction with US partners. We’ll see e-labs within e-learning offerings, as an extra feature within classroom courses, or as an option within blended solutions.

Logilent is developing a lab assessment technology, which will be able to measure not only whether a student carried out the task successfully but whether they chose the optimum path. With this additional capability (not that it will be easy to develop), e-labs could even be used for accreditation. Possibilities also exist for e-labs to be used in recruitment – don’t just take the applicant’s word for it that they’ve got the experience and the qualifications, find out for yourself how they get on when confronted with real-world networking problems. Can’t wait for the new assessment software? No problem, just sit and watch.

In summary, e-labs appear to constitute a genuinely new option for the training of networking specialists. They don’t really threaten the status quo as they are designed to work alongside existing media, whether face-to-face or online, instructor-led or self-paced. E-labs are simple to implement (as long as you’re not the provider), easy to access and relatively inexpensive. Why not try one for yourself and see.

Resources

Providers of e-labs in the UK
Logilent: www.logilent.com;
Global Knowledge: www.globalknowledge.net;
Spring: www.spring.com; 0800-195-6-195