AnyLogic North America

Play with trains at work! The AnyLogic Rail Library

Last post I highlighted the new Traffic Library but it was brought to my attention that the Rail Library barely got a mention in passing.  Let me fix that.  You have had the capability to model trains before in AnyLogic, after all it's a general purpose environment, but the new library makes it drag and drop easy.  You can lay out track or have AnyLogic create a track layout from a descriptive file.  You can have multiple cameras.  You can easily combine trains with pedestrians to get a train station model (see video).  You can import custom cars if you want more than provided in the library.  You can simulate kicking cars or humps.  The trains understand switching logic.  In short, we've shown this to a few people in the train business and they are mighty impressed.

 

To see more videos go here http://www.xjtek.com/anylogic/resources/videos/

But that's not all!  The XJ team has been hard at work on writing the Big Book of AnyLogic and I am happy to announce that the Rails chapter, in all its shiny newness, is available.  You can read it at the XJ website, download it, or just flip through the first 15 pages of it here to get a taste.  Click the right arrow to read this in its frame, click the center to fill your screen.

 T

To run some train simulations on our website go here  Or you can look at some of our newer models on YouTube

Sneak Peak at Coming Road Traffic Library

Anyone who has ever used the Pedestrian or Rail Yard Libraries in AnyLogic knows how powerful they are.  Just drag these objects into your design window from the palette and BANG! ZOOM! they bring a lot pre-build capabilities with them.  You can have pedestrians running around an architectural layout in practically no time.  And they already know how pedestrians behave.  No additional coding to teach them they can't walk through walls, etc.  And for the last few months you've also had the ability to do this with trains.  They're smart trains too; they know how signals work. 

Now, as an exclusive for the AnyLogic North America blog, I can tell you that all this coolness is coming to the May release of AnyLogic in its new Road Traffic Library.  

And of course it comes with a set of existing vehicles (more can be imported), 3D awesomeness, and the ability to leverage all of AnyLogic's extensive features.  Vehicles behave like, well, vehicles.  They know what different traffic signs are and how to react to them.  Just give your little agents the keys, step back, and watch them drive around (all without impacting your insurance premium.)

Have a particularly detailed or complex bit of logic you need to employ?  No worries.  Since AnyLogic is a pure Java app you can easily extend or modify your simulation at any point with the behaviours you are interested in.  Our designers thought of a lot of things, but not necessarily everything.  So with our foundation and your modifications we'll make a great team... with a time-to-complete modeling project that nobody else can touch.  (By the way.  If you DID come up with some neat code or new requirements by all means tell us!  We can highlight you on our web or even include your needs in a future release.  I'd tell you about our upcoming Model Sharing website but I really should keep it to one Leak per blog post.)

If you want to keep tabs on this release, or if you are just new to AnyLogic and want to see what I have been talking about, pay a visit to the XJ Technologies website or, if you want to look at a pedestrian model of what I have been describing you can go here and click on any of these examples.  They'll all run in your browser without any other software needed.

AnyLogic adds telephone support for key accounts

As all AnyLogic users should know our support is AMAZING.  And by "amazing" I mean: unlimited, in-depth, really smart, helpful, built right into the tool and, well, email-based.  But the email part makes sense; the timezone differences and the detailed types of responses required make a quick telephone answer less useful than a carefully developed bit of code or model-testing.

But what about those times when you're just stuck and you need a simple prod to move you forward again?  We are happy to say that in 2011 we will be rolling out telephone-based support, located in the US Central Time Zone and staffed with native English speakers.  For now we are rolling this out to our larger National Accounts as a test -- so don't be surprised that I am not publishing the number here!  We have high hopes that this will be a success and we may roll it out on a broader basis.  Perhaps as a feature in Professional or University Researcher licenses?  Please let us know what you think!

As always, if you have any commercial or process questions you can call the US sales office anytime at (480) ANYLOGIC.  

Facebooks Location Feature Will Help Companies Build Better Models

Facebook is preparing to roll out a new feature that will identify user’s locations and will allow the ability of advertisers to more accurately target users geographic data.

McDonald’s will be the first brand to test the new features. The McDonald’s integration will involve users checking in at McDonald’s restaurants and showing featured food items in their posts. Digital advertising and marketing shops around the country are preparing to construct campaigns around this new functionality.

What this means for Agent Based Simulation:

By integrating the data of some of the 400 million Facebook users, agent based models and simulations will be able to more accurately simulate the dynamics of the market.

Knowing what and where users react to advertising and promotions will allow companies to identify the trends of multiple markets, not only geographically but by age and status as well.

Agent based simulation and system dynamics are only as good as the information provided. The ability to tap into enormous user groups who have a record of willingly providing valuable information into likes and dislikes will prove incredibly valuable.

How Healthcare Simulation is Helping Us Understand Diseases

For much of the last century, modern medicine has invested in the treatment of diseases and illnesses, and very little on the prevention of these diseases.

Part of this is due to the invention of antibiotics and disease fighting medicines that had never been available before. This ability to identify the illness and prescribe a solution to said illness was found to be much easier and much more effective than planning for and preventing diseases. The reason for this is because of the many variables involved. With prevention you are trying to identify the root cause of a problem and identify a plan for preventing it. This was thought to be much more expensive than simply identifying the problem and treating it.

Healthcare Simulation

With recent advances in healthcare simulation we are finding that this is not always the case. Take for example cardiovascular disease, the number 1 worldwide killer of men and women according to the Mayo Clinic. With a simple simulation model, similar to the one that can be viewed here, we can quickly identify the enormous cost of treating this disease, as well as identify those most vulnerable.

By lowering the costs of prevention, healthcare simulation is playing a vital role in our understanding and prevention of diseases.

To learn more about this topic visit our white paper page on healthcare simulation.

What Was The Simulation Analysis for the Oil Rig Collapse in the Gulf of Mexico?

Two weeks ago off the coast of Louisiana a BP oil rig collapsed into the Gulf of Mexico. Since that time we have seen close to 5,000 barrels of oil per day spill into the ocean.

What did BP do to test its technology and did they incorporate simulations of accidents or failures? An article in the Wall Street Journal cited Rep. Bart Stupak (D., Mich.) as saying "'What did you do [before the accident April 20] to make you believe this is really going to work?' We never really got an answer," Mr. Stupak said.

Was a simulation model run to test the impact of the potential collapse and what was the simulation analysis of such an event?  

It is easy to pick apart these events in hindsight and perhaps a more in-depth investigation will reveal what really happened.

This shows how accident simulation is no longer reserved solely for nuclear installations, rather simulation modeling can be used to help organizations predict and react to everyday business processes as well as worst case scenarios.

Read all about AnyLogic 6.5 in our Press Release.

http://www.free-press-release.com/news-anylogic-releases-version-6-5-of-its-leading-simulation-and-modeling-environment-1271617877.html

What color is your parachute? Maybe it's our color.

We're looking for a leader to join our consulting team! 

An exciting opportunity for a motivated self-starter with the appropriate business experience and consulting skills.  AnyLogic North America, while a startup, is a wholly-owned subsidiary of XJ Technologies of St. Petersburg, Russian Federation.  Our primary line of business is marketing and selling AnyLogic- one of the premier simulation modeling applications on the market and the only one with the capability to model in multiple paradigms: system dynamics, agent-based modeling, and discrete event or process modeling.

As a leader of our consulting team your primary responsibility will be to build our North American consulting business.  You'll be given a large degree of freedom including development of marketing strategy, pricing, delivery, and coordinating with our consulting team in St. Petersburg. Required skills are a combination of business development and sales preferably in technical or management consulting and simulation modeling expertise.  Knowledge and experience with our AnyLogic software is, of course, highly desirable and will be heavily weighted in our evaluation. Other responsibilities include: project management of any work outsourced to our St. Petersburg team, participation in trade shows and conferences, delivering face-to-face training, providing prospects with software demos, and providing tech support for US customers requiring either US citizenship or immediate support via telephone.

Compensation is base salary plus significant commission.  Travel requirements, other than trade shows and training, will be determined by you as you execute your approved sales strategy.  Relocation is not necessary as we are comfortable working via web meeting, Skype, and email.  Please send resumes to: George Rivas, Director of AnyLogic North America, via grivas@anylogic.com

How can we cope with all this incredible technological change?

You've all seen the "Future Shock"-like videos and ppt's... the ones that tell you that it took Facebook 2 years to reach a 50 Million user penetration level while it took radio 38 years... and that more unique information will be generated next year than in the previous 5,000 years. Scary stuff like that. (If not, here's a fair example:http://www.youtube.com/watch?v=3cfmS6dmfd0 ) 

So if we find ourselves at the inflection point of some exponential curve re information/complexity how can we possibly cope? We can't say that "every generation has felt this way" because it just isn't true. Growth has not been linear it has been accelerating very rapidly. You have seen more change in normal business practice in the last two decades (desktop computers, internet, cell phones, text messaging, IM, iPod's, GPS, Blackberries, PSP's, Cloud computing, location awareness, iPad's, Second Life, Augmented Reality ) than your grandfather and your great grandfather saw in their entire careers combined. 

The only silver lining that I can see is that our tools have been growing in strength and features too. Your great grandfather (somehow) managed projects by mail and chalkboard and didn't even know what a Gantt chart was (ca. 1915 but not popular until much later). Nowadays you have integrated workflow and collaborative tools that are updated automatically and nearly instantly via smartphone and web meeting. (Ok. Some of you have them.) 

One of the reasons I am in the simulation business is that I believe these tools are the next killer app for scientific and business problem solving. Table-based algebra (aka spreadsheets) had its chance; the market may not have marked Excel for extinction yet but I believe it will diminish in importance and ubiquity. Nowadays we can create a simple agent-based simulation for millions of agents running on a laptop in less time that it takes to create a spreadsheet budget. And these models (unlike deterministic spreadsheets) are rife with probabilistic, stochastic, and emergent behaviors. Experiments that actually TELL YOU SOMETHING as opposed to quantifying your assumptions. 

Processor power, computer languages, and UI have grown to the point where you don't have to be a top scientist to run simulations. What's more, tomorrow's workforce will have grown up playing The Sims, and Civilization, and Age of Empires, etc., and will be quite comfortable with this approach. They will be more receptive to systems thinking, feedback loops, and the like. And while the market is cluttered with small vendors it seems that none of the large software houses have cottoned on to this. Where is Microsoft SimPoint? 

So let's hear what you think. Is simulation an evolutionary dead end in analytics like Artificial Intelligence proved to be or is it the next Social Networking movement and we all just barely have time to get in on the ground floor?

The Simulation Zeitgeist

The simulation/modeling world, like Gaul, is divided into three parts: 

• Systems Dynamics which can trace its beginning to the 50's with MIT's Jay Forrester, 
• Discrete Event which started about a decade later largely with Geoffrey Gordon of IBM and other places, 
• and the newest arrival on the scene Agent Based Modeling.  

These three approaches look at problems in varying ways, are applied with different tools, and have distinct (sometimes determined) followings.

It is no more profitable to discuss which is better than it is for carpenters to argue about the relative merits of hammers versus saws.  For various applications they are each more suited than the other. But since there can be considerable overlap between these modeling approaches, and since most practitioners become adept at only one of them, it is inevitable that comparisons (aka disparagements) are made. Most of them are subjective and qualitative.  In martial arts cinema the "my Kung Fu is better than your Kung Fu" argument is usually resolved in Reel 3 with a lot of grunting and broken crockery.  We have no such simple resolution mechanism.

But perhaps "better" is not the right question?  Perhaps we should be asking which is "hotter"?  Which is on the ascendant?  Where does the future lie?  I have attempted to capture the Zeitgeist (from the German words for time and spirit) -- the prevailing mood and beliefs -- of the simulation community from the available sources.  The pie charts tell a tale. 

1. Total Web Hits [www availability ca. 1990] - a simple Google search show that this is fairly evenly distributed.
2. Downloadable docs [capability common as of ca. 1993] - It is clear that Discrete Event dominates this category.  These are all PPT's or PDF's, sales brochures as well as scholarly treatises.  This may possibly reflect the larger number of software vendors and the long tenure of several of them.
3. Online Forums [capability common ca. 1999] - Switched searching from Google to Zhift so I could focus specifically on live forums.  For some reason System Dynamics is very dominant in this area. Or perhaps it's not surprising that a language specializing in feedback should have so many feedback sites.
4. LinkedIn [launched 5/03] - Searched for members in groups dedicated to these different approaches.  It's clear that ABM is very lightly represented.  Is this because the user base is younger and less interested in business networking?
5. YouTube [launched 2/05] - We see a huge swing towards the ABM approach and very little SDS. Video, as a medium, is just not SDS friendly.  Whereas ABM and DES both emphasize the visual -- the former in analysis (mostly) and the latter in results presentation (mostly).
6. Twitter [launched 3/06] -- ABM again dominates (based on total tweets in the last 5 days) scoring as high as the other two combined.  It doesn't take a demographic genius to sense that ABM is tilted towards a younger crowd.
7. SlideShare [launched 10/06] -- this is interesting.  DES is low but SDS is high.  I suspect this is the medium again.  SlideShare is Ajax/Flash automation of ppt's (mostly) and doesn't require actual videos like YouTube.  It's a great place to breathe new life into more or less static slide decks.

Is there a discernible trend?

The overall volume of web hits and web documents must, at least in part, be attributable to the fact that SDS and DES have just been around a lot longer.  When I look at the demographic bubble implicit in the YouTube and Twitter statistics I would have to conclude that the time of the Agent Based Model is coming.  

What do you think?  Give us your comments.