"The Great ‘Paramics’ War"
Microsimulation Forum’s guest columnist Richard Braidwood, from Braidwood Associates, dons his tin helmet and goes behind enemy lines to find out the latest on the great Paramics war.
Introduction
It is perhaps ironic that battle lines are once again being drawn up on the 1st of August this year – 91 years after the start of the Great War – between two neighbours who, in recent times, were allies but are now battling for global domination. August 1st this year saw the first advance across territorial boundaries as the Great ‘Paramics’ War ensued.
Background
Paramics is an acronym for Parallel Microscopic Simulation and was first used in 1992 on the University of Edinburgh's Thinking Machines CM-200 and later on a Cray T3E, then the fastest civilian supercomputer in the UK. At the time, these were the only machines powerful enough to cope with the enormous amount of processing required. The rapid increase in the power of a desktop PC over the subsequent 5 years resulted in parallelism being dropped from the released product by 1997.
In 1995 several of the University's staff spun out a privately owned technology company, Quadstone Limited., and brought with them the prototype Paramics software. Soon after Quadstone was founded, they joined with SIAS, a transportation consultancy founded in 1974, to form a joint venture company, Paramics Traffic Simulation Limited. SIAS had been one of the original collaborators on the Paramics project. The aim of the joint venture was to develop and promote microsimulation, but it was dissolved in early 1998 when Quadstone and SIAS failed to agree on the direction of the company.
Following this split, a lengthy treaty was signed after exhaustive negotiations between the former allies. This meant SIAS were restricted from selling in the Americas and Japan and Quadstone were restricted from selling in UK and Ireland. But crucially each had their own product to sell, initially identical, but gradually diverging.
Which was the real Paramics? Whose idea was it anyway? Did the end users really care, if the product worked? One of the Quadstone founders, Gordon Duncan, now director of Azalient Ltd, is recognised as having written most of the core Paramics software, but does not lay claim to being the original source of the idea of Paramics. Gordon Duncan comments "I don't think the provenance is that important. Many people over the years had a similar idea. There must be almost as many half-finished, half-working traffic models as there were traffic modellers. The difference with Paramics is that we made it work."
The stand-off ended on 1st August 2005 when the treaty expired. Geographic boundaries no longer restrict both companies' business advances. Is this Glasnost? Or are we in for a Battle of Britain?
Paramics the Name?
One outstanding element of the agreement was product naming. Neither party would agree to the other owning the name, which even in those early stages was a recognised global brand to modellers in the know.
At this time the product name was not perceived as critical as each company had its own defined market and this short-sightedness has led to the confusion we now have – perhaps neither company considered the successful uptake of the product.
So we have ‘Paramics’, Q-Paramics or Quadstone Paramics and on the other side S-Paramics or SIAS Paramics- confused, many people are, and no wonder!
The name Paramics is effectively redundant given neither version of the software uses parallel processing. Is this an opportunity for one of the companies to change their product name to differentiate their product offering?
Since SIAS recently rebranded their product to S-Paramics we will for the purpose of this article call the SIAS version S-Paramics and the Quadstone version Q-Paramics.
So why won’t they rebrand? We all know the benefits of generic naming and the strong brand associations attached to products. The associations with Paramics stand for security and authority to buyers however the name itself is dated and doesn’t do what is says on the tin. We have no way of knowing who’ll do what but someone needs to make the change from Hoover to Dyson soon in order to trail blaze this software.
What is for certain is that this extraordinary situation of two companies marketing two different products with the same name in the same marketplace, being developed in offices only a matter of meters apart is sure to intrigue those in the traffic and transportation sector.
Software development – which company is better?
Quadstone is a software company with a traffic modelling team.
SIAS is a transportation consultancy with a software development team.
Which formula is better?
Ultimately the projects and users drive software development in some form whether it’s a feature development for a particular project or multiple users independently requesting the same feature.
’The Customer is Always Right’ should be applied in any competitive market and the Microsimulation traffic model development field is no different. There are always exceptions, feature requests have to be within reason, a model calibration button is a little too much to ask, and could have implications on the need for our traffic modelling skills! The answer is, the formula is effectively the same and both products are being developed in their own ways - driven by the markets they dominate.
Should the greater number of licenses in circulation and users globally make for a better more rounded product? If you have a larger global user base this should influence the depth of the product and increase the ability to source minor yet sometimes frustrating bugs. Conversations with Quadstones Richard Millington and SIAS’s Stephen Druitt about user numbers produced quite conflicting reactions. Quadstone’s, Richard Millington, General Manager at Quadstone Paramics commented 'with over 2000 users in 40 countries across the globe, I feel confident in saying the Quadstone Paramics software is the most comprehensive micro simulation tool available today. We continue to sign new customers from our core territories in North American and Australia which reinforces our position as a market leader' Although it is safe to assume that SIAS dominate the UK marketplace, Stephen Druitt adopted a more cautious approach preferring not to disclose SIAS Paramics user numbers.
So is this a battle of ‘bravado’ from Quadstone an army on the attack and SIAS, the defensive tactical platoon not leaving the bunker to expose itself to the line of fire? Should Quadstone be worried about potential advances in the US from SIAS and is this distracting Quadstone and SIAS from other threats, Aimsun and Vissim? Are their armies on the march? A story for another article perhaps….
With an extensive user base in the United States, Q-Paramics is exposed to arguably the most congested freeway and urban road networks in the world. This makes the Q-Paramics application in the UK market potentially much easier as freeway/highway congestion and traffic volumes are significantly less. Yet S-Paramics has been developed to the needs of the UK market therefore Q-Paramics may not have the features that users have been used to in the last few releases of S-Paramics. Having used the latest versions of both products, they ‘could’ be applied anywhere in the world with some changes that can be identified from market knowledge.
Having been involved in development, trained users and applied both products in live projects, in the UK and throughout the world during the last 9 years I can say they both have positive and negative features much like the Microsoft products developed by Mr Gates.
Which should you purchase?
The million dollar question, everyone has preferences, BMW, Volvo, or Vauxhall and of course there are other microsimulation traffic modelling tools available on the market, Vissim developed by German company PTV and Aimsun developed by Barcelona based TSS.
Now that either Paramics product can be purchased anywhere in the world, which should you choose?
Restrictions on sales in certain regions have prevented many modellers getting the opportunity to use both versions of Paramics so not many people have knowledge of the differences, least SIAS or Quadstone. If you talk to either company you will hear something along the lines of ‘we are the best’ said with enthusiasm and anger normally attributed to a tribalistic football fan. A handful of ex-SIAS staff who have ‘crossed those boundaries’ including myself have been involved using both tools and have the knowledge of applying both products in any desired capacity, a strength for sure and the comfort of not concerning ourselves about the impact if this war is to result in fatality. Nor do we have anything to benefit from by promoting one product over the other.
What does the future hold?
The thought of most who don’t know the background is, ‘would they not be a stronger force together’, almost definitely, the next question is ‘Will they ever work together?’
SIAS and Quadstone are unlikely to call a truce on the ‘P’ war therefore a relationship where they are allies is doubtful as agreement on common leadership and direction would be required to co-ordinate their forces activities, something that past issues would prevent…..‘never say never’ I suppose.
Conclusion
What ‘Paramics’ has brought to the world traffic modelling market is the methodologies of calibration and validation of traffic models. Industry standard analysis used to compare modelled and observed turn and link flows, the GEH statistic (the goodness of fit statistic defined in the Design Manual for Roads and Bridges), is now commonly referred to in many US standards such as the Federal Highways microsimulation guidelines which many DOT’s have adopted within their own guidelines.
Collectively SIAS Limited and Quadstone Limited should be congratulated for being at the forefront of creating a product that has had a major impact on the global market of traffic modelling, one of the few things that I suspect they would ever complement each other about.
A more general view on a major issue facing the microsimulation traffic modelling market is the quality of the application of all the microsimulation products, inexperienced users or those without a satisfactorily knowledge of the products, through no fault of their own can mean poor project results reflecting badly on the tools and create suspicion in whether this really is the way forward… which inevitably impacts market confidence.
The key differentiator between the Paramics products will therefore be the accuracy of the software, the ease of use, the number of people using the software (VHS vs Beta – because the US market are unlikely put up with two versions of the product) and most importantly, customer service. |
发表于 2008-7-31 23:08:09
Post By：2008-5-30 10:36:00 
发表于 2008-8-1 09:57:21
楼主