COFES 2005

2928 0

Each year in the desert of Arizona, the Computer Aided Design community gathers to talk about the future of engineering software at Cyon Research’s unique event. Martyn Day was in attendance.

The CAD industry has many annual conferences and events; from specific vendor’s shin-digs like Autodesk University to those with vertical market focus, e.g. Daratech Plant. Cyon Research’s The Congresss on the Future of Engineering Software, or COFES is something quite different and has been created by the industry research firm, Cyon Research. Held in Scottsdale Arizona, the event attracts representatives from the majority of Engineering software vendors, together with industry notables, customers, analysts, start-ups and business angels. While the agenda is fairly mixed and split between Mechanical CAD and Architectural CAD, there is plenty of time left over for networking and ad hoc technology demonstrations.

This years event was entitled, “Innovation in an Idea Economy: Putting Your Money Where Your Mind Is”. Recognising that new products are usually applications of existing inventions, brand new ideas and concepts come along less frequently but these spawn new industries and technologies. Cyon explains that 100 years ago it was the car; 50 years ago, the transistor; ten years ago it was the Internet, so what’s in the pipeline? With so many manufacturing jobs leaving the Western economies, Cyon wanted to push the concept that wealth will shift to those industries and those economies that foster inventiveness and innovation, meaning the high-paying jobs will exist in those economies that recognise the value of knowledge creation and capture, the fundamental building blocks in a knowledge-based economy. According to Cyon, it’s no longer enough to build products better, cheaper or faster. The winning hand goes to the player that is able to innovate and capitalise on its intellectual property. It’s not just about what tools you use to design, it’s about how you use, analyse and reuse that data.

While other such forums have all been sucked into the bottomless pit of PLM verbiage and marketing, Cyon has managed to keep COFES remarkably free from the warping powers of the really large vendors, yet they still send representatives to listen and join in the mix. That’s not to say PLM isn’t a topic for debate but it’s just not ‘all dominating’.


The arrival day started off easy with a late afternoon talk given by Peter Thorne, of UK analyst, Cambashi. Peter gave a great presentation explaining to the many Americans at the event, what was happening in Europe, compared to the US. Being based in the UK, it is quite amazing what many US-based companies think is going on in Europe. There still seems to be a ‘one size fits all’ attitude to Europe – that and thinking what happens in the US happens in Europe, eventually. The key take-away point was that all the new EEC countries are seeing most of the growth as investment flows into these low-cost manufacturing economies.

Day Two started with Dr David Weinberger’s analysis of the web and how information is structured. David sees the web as a key instigator in freeing information from previous classification and storage methodologies, where data is grouped and stored hierarchically in categories. Web search tools and the nature of the web’s intrinsic messiness, means unstructured data is good and free from limitations. Many examples were given, especially with RSS feeds and readers but the underlying message was that knowledge and authority over knowledge will change.

Day Three introduced Peter Marks, of Design Insight, asked what the future of engineering software was in an economy where manufacturing is in decline and services are on the rise? One of his conclusions was that the services vs. manufacturing debate is largely irrelevant. Services merely shift the focus and intensity of new product development and manufacturing from end users to service providers. Another conclusion is that a new breed of engineering software will emerge – one that better supports services. For software and hardware vendors this means a shift in both target customers and software offerings. For users, it means reevaluating what it means to innovate. In previous talks Peter predicted the non-event of the Y2K issue and also predicted the bubble in the dotcom growth


The main homage to PLM took the shape of a panel, chaired by Dave Burdick of Collaborative Visions. Entitled, ‘Fuelling the next engineering market growth cycle’, the panel included: Buzz Kross (Autodesk), Brian Shepherd (PTC), Raj Khoshoo (UGS) and Dominique Florack (Dassault Systemes). PLM has been relatively stagnant, growing at 5% or less per year and still being 1/3 less in revenue size than ERP. It has under-performed its market predictions, so how can growth be generated? This promised to be quite interesting but ended up with each vendor just trying to justify why each of their respected visions of PLM was more meaningful. PTC appears to have lowered its sights from the very high-end to the mainstream and small business workgroups. Autodesk is just about delivering a very basic EDM system, with pretensions to get a PDM product in place over the next three years – however they have huge numbers of users and a low cost of entry. Raj from UGS was probably the most comfortable talking about the high-end as TeamCenter has undoubtedly been the biggest success from UGS’s merger with SDRC. Dominique from Dassault talked with great passion about Dassault’s vision and probably was the most compelling, as the Catia/ morphing vision being deployed by Toyota sounds so space age, however Enovia and Delmia sales have yet to come to the fore at the company, with SmarTeam getting more success at the PDM level. The fundamental fact is that PLM has yet to gain any traction with users outside of Aerospace and Automotive – after millions of dollars spent on marketing and talking about PLM, the majority of customers I talk to still don’t know what PLM does, unless they think it’s PDM. The industry analysts are responsible for over-hyping PLM for their masters that it may never live up to the expectations that were originally associated to it.

There is a limit to how much change you can inflict on an installed base and while many innovations are welcome, there are many, many more that are not adopted

Publishing Formats

I’ve written about the engineering ‘Publishing format’ war that’s currently going on many times and it just so happened that the first ad hoc meeting was with one of the independent players, Lattice 3D. The publishing format issue provided me with a running thread of demos, conversations and bemusement throughout the three days of COFES. Off the top of my head, the congress attracted: Adobe (PDF), Dassault (3DXML), Autodesk (DWF), UGS (JT), SolidWorks (e-drawings), Lattice 3D (XVL), Actify (.3d), Intel (U3D), N-Grain (voxel-grid based) and Right Hemisphere. I would hazard a guess that we have more proprietary 3D publishing standards than we have actual proprietary CAD formats. Our industry strikes again – we all agree that standards are a good thing, indeed like tooth brushes, but nobody wants to use anyone else’s. One could argue that this somehow provides choice for the industry; it unfortunately just means possessing many free viewers on your PC.

My main exposure to the publishing formats on this trip were with Adobe, Lattice 3D ( and N-Grain ( My first meeting was an impromptu demonstration of Lattice 3D. Lattice is originally a Japanese company that ‘s partly owned by Toyota and in fact Toyota uses it internally. Lattice 3D is a reseller and developer of the product – this caused some confusion and I will come to that later. The system is without question the best format for storing 3D data that I have seen to date. Based on the XVL format, Lattice has worked out a way to describe solids as a series of NURBS surfaces, this allows incredible control of the size of the XVL file but also allows scaling and some degree of control as to how accurate the portable file can be viewed or printed. The competition tends to break down models into just tessellated surfaces, which can be a bad approximation. UGS’s JT format sores a number of different representations at varying resolutions of tessellation. As Lattice went for NURBS, it’s very much like Postscript – highly scalable from just the one source file. It can be incorporated into Office documents, marked-up or displayed streamed over the web.

It’s not just popular with customers, Dassault system uses the technology as a key component of its 3DXML publishing format and there are rumours of other vendors licensing the products capabilities. As the event had both Dassault and Lattice 3D present, at times I felt like I was the ping in the pong between the two companies, while trying to get to the bottom of who did what and how Dassault works with lattice, not Lattice 3D. Still, when you see the technology it’s just obvious why it has advantages over most of the other 3D formats out there today.

N-Grain is a company that’s new to me but it’s also a different take on the old publishing format. The company uses a Voxel-grid engine to represent the geometry. A Voxel grid is a ‘gappy’ representation of cubed shaped grids, almost like a 3D raster model. Depending on the resolution as you zoom into the model you will see some rasterisation and ‘zaggies’. Still the models are pretty small and easy to manipulate.

N-Grain doesn’t have any large partners like Dassault to help push the technology and so has been left to forge ahead on its own. To date the company has focussed on the military market and specialised in the creation of manuals and websites with 3D interactive graphics. I’ll be looking at the N-Grain technology in more depth next month.

Adobe ( was at the show in great numbers, emphasising the sheer size of PDF in the Engineering market. While I think Adobe made an error selecting Intel’s U3D format, as nobody else supported it, it seems that Adobe is mustering many more vendors to start supporting U3D output and has made some strategic technology purchases to enable the capture of OpenGL 3D data to create U3D models – this bypasses the need to get at the data through all the different CAD systems’ APIs. Autodesk is also developing similar technology for its DWF format.

The one technology from Adobe that is attracting my attention is the Policy Server technology that an organisation can purchase to control documents in circulation without a typical document management system. At the moment it’s not cheap (about $40K) but it appears to offer unprecedented control over important documents. PDFs that are emailed out can check back when used on a system that’s on-line (you can chose for the PDF to be only viewable or not when there is no Internet access). Using the policy Server you can control the level of printing (high, low or no printing), immediately stop access, set a cut-off access time, count the number of times it has been opened, see if anything has been cut and paste out of it or inform the user that a more up to date version of the document is available and point out where it can be downloaded from. Turn to page 22 for more on this technology.


So many people, so little time! While the morning sessions are usually kicked off with a presentation, the afternoons are reserved for private meetings and visits to the various companies’ technology suites. On the first day the industry analysts are on-hand to give their interpretations of what’s happening, while the second is more technology and development focussed.

Here’s a brief run down of some of the people I met, what we discussed and some cool technology that’s coming our way in the next year.

Bob McNeel, CEO, McNeel Associates

Bob McNeel is one of the most pleasant people in the industry and COFES gave me an opportunity to catch up with what’s been going on at McNeel Associates. McNeel develops and sells Rhino, the low-cost surface modelling tool which has over 150,000 customers worldwide. The software is used in all sorts of industries, from footwear to architecture and is certainly excellent value for money. Bob won a lifetime achievement award at this years’ COFES and it was well deserved, as Bob has provided great leadership in providing value with his CAD dealership and innovative software development (with products like AccuRender). I asked Bob if he would ever contemplate selling out to a big player, to which he replied after some contemplation, “is there enough money out there for someone to give me to stop doing what I love doing? Probably not.”

Our conversation then went on to today’s innovative products, Bob having just come from an optics conference. There he had seen a laser device that could work out diabetic’s glucose levels without the need for needles, a scanning electron microsope that fits in your hand and the military showed a laser that could cause burning sensation in skin, without causing any damage (one wonders what the last device would be used for?).

Jon Hirschtick

Jon is an industry giant, and I’m not just referring to his height, then again everyone is taller than me anyway. John was one of the founders of SolidWorks and is a regular COFES attendee and while he’s there to represent SolidWorks he is also known as a business angel and investor. Throughout the event many of the new start-up firms would seek out Jon’s opinion on their product and look for advice.

We talked about a number of issues in the industry and talked beyond CAD into the realms of open software, an area which John is particularly excited by. The quality of free software that’s available for Linux has amazed him. On the CAD side we talked about the complexity of SolidWorks (the product) and the increasing difficulties of improving software quality as these CAD products increase in size, complexity and functionality. Jon acknowledged that SolidWorks had issues in the past with bugs and stated that quality was now a key focus within development, starting with the forthcoming release of SoldiWorks 2006. We talked a little about ‘Cosmic Blobs’ and how the company was looking to innovate to make modelling easier, yet more powerful. While Cosmic Blobs is essentially a kids toy, if follows the Xerox Parc methodology that if a kid can use something, then an adult can – development that ultimately led to the Windows interface.

SolidWorks and Autodesk are still locked in combat for the mid-range CAD market. SolidWorks has developed eDrawings for Inventor, provided a DWG compatible 2D editing tool and recently added DWG gateway – a potentially powerful tool to save AutoCAD DWG files in current or past DWG formats. It’s pretty clear that SolidWorks is trying everything it can to get into those 2D AutoCAD accounts. Jon and I talked about the strong bond that SolidWorks maintains with its customer-base, which is unique in the industry, despite hundreds of thousands of users. While SolidWorks is keen to grow towards its target of a million customers, at the same time it’s looking at ways to keep its relationship with those users. It’s a tough task. The market also seems to have stronger competition these days, with PTC on the rebound and Autodesk improving its products and channel.

Brad Schell, CEO, @last Software (SketchUp)

We are big fans of SketchUp here in the office. It’s a low-cost 3D conceptual modelling tool that comes from a small operation running out of Boulder Colorado. In fact the company’s HQ is a shop on the main street of Boulder, complete with large glass front window. I’d never met Brad before but he really impressed me and made me laugh. The product is growing world-wide and even competitors like Autodesk have nice things to say about the product and the development team. The conceptual market has really been a tough nut to crack and SketchUp has been the most successful application in that space. Brad and I talked about how tough it is to make a new market for yourself where all others had failed. The worry being that by showing the way forward, it’s possible for the big vendors with deep pockets to come in later, after the hard work has been done and copy the formula. As things stand, the company is still adding considerable new functionality to each release and shows little sign of slowing down. Hopefully with a growing base, an innovative edge and a good view of customer relations, @Last will either carry on making its niche, or be made an offer they can’t refuse from one of the industry giants.

If you haven’t seen Sketch-Up yet, go to the website and download a free trial.

Arnold Van der Weide President IntelliCAD Technology Consortium

The IntelliCAD consortium started out as Visio IntelliCAD – the first AutoCAD clone which tried to take on the power of Autodesk and provide AutoCAD feature by feature compatibility at a vastly reduced price. However, the product was released too early and Autodesk fended it off successfully with LT and a negative advertising campaign informing users that its products were 100% DWG compatible, not like this AutoCAD emulator.

After Microsoft bought Visio, IntelliCAD found a new lease of life as a product developed by a consortium of Autodesk competitors, acting as a repository for DWG knowledge, providing libraries to develop products like SolidWorks’ DWG editor and DWG gateway. IntelliCAD is also a stand alone CAD platform. Arnold is an industry veteran having worked for many of the major CAD vendors. Arnold explained that he sees IntelliCAD now as being a serious development platform for application specific developers who need a CAD tool, with AutoCAD compatibility, to produce dedicated solutions. Apparently in Holland IntelliCAD forms the basis of the industry-leading tool which is used in dredging the country’s many waterways.

Arnold’s future vision for IntelliCAD seems to divert from the AutoCAD emulator that it was originally created for – adding new features that aren’t necessarily in AutoCAD and not necessarily copying features that are added to AutoCAD, while maintaining a strong support for DWG. This is an interesting departure, although I do wonder what other members of the consortium think of this concept. Talking with Arnold it’s also interesting to hear that the development work is carried out all over the world, in a true ‘virtual’ way.

One has to wonder if anyone will seriously challenge Autodesk’s dominance of the 2D CAD market, especially as the price of LT is heading for the ú1,000 a seat mark? Most of the other vendors are off developing 3D solutions, forgetting that the majority of work is still done in AutoCAD and LT. There has to be space for a decent 2D tool with a brand-name people can trust.

Hewlett Packard

There were a number of folks from HP at the event showing the lastest in AMD Opteron workstations (see The one interesting snippet that I picked up was that Microsoft was paying HP to go around the world and buy-back all the Itanium-based workstations that it had sold. In the US this only numbered 150 or so. These workstations were Intel’s first attempt at a 64-bit processor. Unfortunately it didn’t run existing 32-bit applications very well/ fast and was blasted away by AMD’s latest generation of 64-bit processors. It seems Microsoft didn’t want to continue developing the Itanium version of Windows and so was helping take them off the market, and HP were replacing them with AMD-based workstations. HP played a major part in the development of the Itranium, providing a lot of RISC know-how to the Intel development team. Unfortunately Itanium chips were also very expensive.

‘Technologists’ are insatiable technology addicts and appear to have severe problems coming to terms with why other people ‘don’t get it’

Rick Stavanja

Rick is the technical guru and editor in chief of CADwire, a popular link for CAD news, commentary, reviews, events and articles. Rick was demonstrating a new CAD-related search engine that he was developing. It will be the CAD equivalent for Google, providing a dedicated search tool that specifically scans known CAD resources for more accurate results. The search tool can also be embedded in other sites, as you can with Google.

Ping Fu, President and CEO, Raindrop Geomagic

COFES was attended by many of the industry’s leading lights and while winding down over a few beers I met Ping Fu, who, from everything I’ve read about her since, is well deserving of that title. What’s even more amazing is Ping’s life story, from ‘leaving’ China for offending the authorities for writing about human rights issues, to starting from scratch in the USA, going through college, finding work, achieving success and then taking huge risk in turning around he point-cloud firm which she now runs. Hearing this story is a very humbling experience, to a ‘lardy’ westerner that’s never really had to work hard for much.

Wilfred Grabert, GiveMePower

On the final day, over breakfast, I talked with Wilfred Grabert, CEO of ‘GivemePower’. Wilfred used to own some AutoCAD dealerships in Germany and the UK but now he has started a company with his son and developed some innovative AEC software. The concept behind the solution is that it’s difficult to capture and digitise the internals of already built structures. To solve this, Wilfred’s software is a new CAD system which is happy to run on portables as small as PDAs or on normal workstations that work in conjunction with the current generation of GPS and laser scanning measuring tools. By walking around the building and taking measurements, the system automatically builds up a highly accurate 3D model of the spaces. This could revolutionise refurbishment work and could be used to check as-built against architects plans.


On reflection, while I write about technology and still run the applications, COFES is both a heaven and a hell in the same place, at the same time. The event attracts some incredibly smart people, individuals who have actually made this industry possible. But there is also a negative of getting so many technologists and futurists in one place. By the end of the event I was convinced that everyone was on some kind of ‘Technology Prozac’. There is a reality to technology adoption and it’s a fact that we can innovate and regenerate products and ideas much quicker than we can understand them and deploy them. There is a limit to how much change you can inflict on an installed base and while many innovations are welcome, there are many, many more that are not adopted. Selling CAD tools on their productivity benefits may no longer be a valid reason, as we have over capacity to produce in so many industries. In some ways this is probably why PLM has been the mantra for so many developers as it expands their influence and revenue possibilities outside of the engineering department.

Jumping to the summing up sessions of both AEC and MCAD, they went along the lines of ‘We have this great technology, so what’s stopping wide-spread adoption?’. While many reasons were given, there is little understanding of the psychological impact of constant change and how customers adopt technology. While we have perfected the art of development and incremental updates in Engineering, we are but flesh and bone and have internal capacity limits. ‘Technologists’ are insatiable technology addicts and appear to have severe problems coming to terms with why other people ‘don’t get it’. Don’t get me wrong, I’m really glad these people are here, always looking forward, identifying trends and looking to solve both current and future problems but reality should dawn once in a while. Maybe my net contribution to the planet will be discovering a pessimism pill? (obviously to be taken with a glass of water, half empty).