Cadtest, from the Bristol-based company of the same name, is a relatively new automated assessment tool for measuring people's AutoCAD skills.
The results are pooled (individuals' details are protected) to form a national benchmark, which is published in collaboration with the AJ.
The application depersonalises the process of skill assessment and, important for those hiring staff, it conforms to a national benchmark. This gives the user the unique ability to compare like with like: you can establish with a degree of exactness where a person being interviewed can fit into your organisation. In addition, you can compare the average result of all your employees with your competitors' results.
So Cadtest is of potential benefit to anyone who works with AutoCAD or works with people who are using it. You can have a confidential, up-to-date picture of how your organisation compares to the rest. As an agency you could respond to your clients' needs more accurately by placing the right people in the right posts. As an employer you could determine the AutoCAD skill level of your current and potential staff. As an employee you might well want to include your result in your CV.
Cadtest can run on any internet-connected computer running a full release of AutoCAD 2002 or later. An LT version is to be released soon. It is compact (only 2.5MB) and so can be downloaded rapidly after you have registered and paid online. It outputs a comprehensive range of results, which are presented in a straightforward way.
Because it is so accessible, it is easy for CAD managers to introduce it to a practice and keep it as part of a company's procedures.
It all sounds so good that the only wonder is that it was not created long ago. I was interested to discover whether these potential benefits made any sense for Fletcher Priest Architects when I gave the software a test drive. I should say immediately that, whatever detailed comments follow, the people who have created Cadtest have got it right on most points.
Starting up Installation is as quick and easy as advertised. Once you run the software you, as administrator, have a few setup options with help screens to guide you. One of the options you have is to take a preliminary demo test. You can also do this now directly from the site at www. cadtest. com once you have registered. I would recommend that administrators take the test several times to get the feel of the software, because it is not as logical as you might wish - and you might want a non-disclosable check on how your own skills are going to be evaluated.
The scoring process relies on you following the instructions to the letter. It will reduce your score disproportionably for quite minor omissions. For instance, at the beginning of the demo test you are asked to create a block named 'MEETING'.
Even if it is correct in every other way, if you name it using lowercase your total score for the 'Blocks' part of the test will drop to zero per cent. This is baffling because AutoCAD is only partly case-sensitive:
it will remember and display symbol names the way you type them but it will not allow you to have 'MEETING' and 'meeting' as two different blocks.
Another part of the first demo test is 'Layer Control'. You are asked to do basic things with layers such as drawing entities on a specific layer, making sure a layer is displayed and what colour it is, and leaving a specific layer current as you finish the test. I was surprised to see that if you do everything else correctly but leave a different layer current, your score for the layers part of the test drops to zero per cent again.
I suppose that the demo test is there for the purpose of explaining how scoring works in real tests, but these zero percentages left me wondering if the national benchmark is skewed because not everyone believes that there is any importance in leaving a particular layer current when they close a drawing.
Armed with the knowledge that every aspect of instructions is vitally important for the final result, I took the real test. The test doesn't interfere with your personal AutoCAD configuration, so you are working in your own familiar environment. Everything ran smoothly and fast, even on my not-sofast computer. The test is broken into 10 sections including lines, xrefs, texts, blocks, dimensions, integration, layers, variables and user coordinate systems, each one lasting only a few minutes.
There are only a few lines of instructions and there is a picture of how the result should look in a small window in the corner of your screen. It was all laid out nicely and very straightforwardly.
I am not an AutoCAD super jock but at the end of the test I couldn't help thinking how easy it had been. Building-industry professionals across the country are using CAD tools that allow online collaboration, smart objects for easy facilities management and threedimensional building models. Having skills in using such tools is the current industry standard. Cadtest doesn't touch on any of those. It concentrates only on the basic level of CAD use. If its tests were wider ranging, and if they covered more of the current trends, you could use it to evaluate your position or customise your training requirements a bit more accurately. As it stands now, all it lets you know is whether you (or your staff) know how to use 10 per cent of the most basic AutoCAD commands.
The people who made Cadtest are breaking new ground and this is a work in progress - they have already got to release 4.1. At this stage the software probably needs some finetuning and a broader scope. But isn't that what we say about all software?
As it is, Cadtest is already good and I believe it will only get better.
Pero Maticevic is practice director of Fletcher Priest Architects. Cadtest costs from £85 for one test down to £60 when you buy 100 tests