LOS ANGELES — Director Gabor Csupo and screenwriter/producer David Paterson are not your typical LA filmmakers. One, along with Arlene Klasky, is the co-founder of Klasky Csupo, here, the production house that specializes in traditional animation–and broke box office records with its Rugrats movies in addition to numerous TV series. Paterson is an indie filmmaker at heart (Love, Ludlow ‘05) and a family man living in a New York suburb.

Neither man had worked on a major live-action motion picture before Bridge to Terabithia and neither would expect to be plunked down in New Zealand with a film crew, a handful of American actors, two American school buses and, situated on the south island, all the digital talent that Weta FX could provide this new film. Paterson co-wrote the screenplay with Jeff Stockwell. His mother, Katherine Paterson, wrote the classic children’s book of the same name in 1975. But it took her son 15 years to bring this production before the cameras, and one major factor Paterson points to is production company Walden Media. (Disney is distributing Terabithia.) Paterson credits Walden (Narnia, Charlotte’s Web and more) as one of the few companies dedicated to adapting classic children’s novels into high quality feature films that remain true to the original stories. Walden also succeeds by marketing its films to a novel’s potent core audience: young readers and reading teachers.
Advertisement

But to succeed, you would need Weta, fresh off its major work on King Kong, to take on this smaller film (at roughly one-tenth the Kong budget). And Weta did: the shop’s VFX supervisor, Matt Aitken, was charged with (”only”) 130 FX shots. And you would need that great, revered story, thanks to Ms. Paterson, to hold viewer interest. The film’s conservative number of VFX shots leaves lots of room for storytelling and, say Paterson and Csupo, preview audiences agreed that the new film offers just the right balance of visual wizardry.

A ROPE & IMAGINATION

The film’s premise follows the book fairly closely: a fifth-grade boy and girl, both somewhat outsiders, become fast friends. They discover a hidden forest possessing magical properties–especially when one’s imagination is applied. Claiming the realm their own the youngsters name it “Terabithia.” The forest, accessible only by a rope swing, becomes their haven as well as the setting for imaginary battles in which the friends use fantastical means to vanquish fantastical enemies. Hence the visual effects.

“In the novel there are references to battling giants and ogres and such, but they were never really assigned names or personalities,” Paterson says. “What I’m very proud of is, although we did create a Terabithia which did not exist in the book, we still followed the truth and points of the novel.”

For the film, threatening Terabithian creatures are imaginative recreations of bullies at school. Stockwell came up with the flying insect army and the fierce squirrel/ogres. Paterson developed the giant character who bears a resemblance to a female school bully.

LIVE ACTION & CG

Csupo got involved with the film, not to become a big live-action/VFX director but “because of the beautiful story. Live action? Animation? What does it all come down to?” asks Csupo rhetorically. “Do you have good characters and are you able to tell a good story to the audience.”

Compositing layers is “not entirely new” to Csupo, “but the CGI with live action was a nice change for me as far as the budget and quality control. Weta was fantastic to work with: very dedicated, very talented, they have all the tools, and it was a blast.” Csupo filmed Terabithia in 35mm with DP Michael Chapman from January to April of last year.

“Simultaneously, Matt Aitken was already on the set,” Csupo says. “We were planning all the visual effects and camera moves. We storyboarded all the Terabithian sequences ahead of time. We shot a bunch of empty plates for placement of the creatures and POV shots of the kids on the set.”

Aitken used a “silver ball”–a reflective chrome ball–on location to capture environmental lighting, Csupo says, “so they could render the elements accordingly to the light conditions so it would totally match the background plates we shot with a live camera.” Another ball–a gray one–helps give exact lighting positions and shadow areas. “He’d hold that up at the end of every take and use it as a reference for lighting when creating the creatures.” Back at Weta FX, Aitken’s people rendered these balls in a CGI environment to match on-location shadows and light-source conditions exactl

Researchers present a modification of a conventional feed water three-element control strategy to solve problems of drum level control

Research studies show there is interest in solving the instability problem of level control in power plant steam generators by designing drum level phenomena models to establish a control law for a feed water control system. However, a model-based control strategy may be difficult for control engineers to understand, especially those with skills acquired in field experiences with proportional plus integral plus derivative controller, or PID, loops but with no mastering in modern control strategies. Now there is a modified conventional feed water three-element control strategy that is easy to understand and solves some problematic situations of drum level control. The new strategy can lead to better performance, especially at full load (generated electric energy) with the supplementary fire (duct burners or after burners) in operation where the feed water control valve is near its full opening. Such a strategy avoids, as much as possible, the saturation of the control signal to the valve, eliminating the permanent oscillation of the controlled variable (drum level). The control system behavior also shows good stability from start-up to full load, including the transition from one-to-three elements operating mode and vice versa, as well as the duct burners entry. It shows similar stability during the stop of the heat recovery steam generator (HRSG) from full load with duct fire in operation to the HRSG out of service.
Advertisement

Due to the software blocks communication system of the maker, a transition band of the steam flow signal as transition signal assures a three-to-one or one-to-three element bump less transfer operation. Laboratory tests simulated the proposed control strategy with a simplified dynamic process model before putting the control system into operation on site. The tests will show real process runs and control-tune parameters.

The case of Dos Bocas Veracruz

The feed water system of the HRSG of the combined cycle power plant at Dos Bocas Veracruz included a conventional three-element feed water control system operating with electric analog components. A vertical steam drum makes the level control with less capacitance (volume per unit of level), and then the level dynamics are more sensitive than that of the horizontal drums.

Modernizing the control equipment by substituting the initially installed analog equipment with a digital one made it necessary to translate the control strategy into a new digital environment. The available tools in the digital medium allowed us to program a more complicated strategy to improve the control performance. However, we need now to deal with time execution of fieldbus elements that introduce dead time due to time consumption and its synchronization time of each element.

Controlling drum level

The problem of controlling the drum level involves the operating conditions of the drum and the deaerator from start-up to full load. Since there is another HRSG connected by a main header, the influences on each other add an additional element to account for the control strategy and tuning.

Due to the rangeability problem in pressure drop, measurement elements in the feed water, and steam flows, we traditionally designed the control strategy to operate at the start-up of the HRSG in the mode of one element, that is, with the drum level signal as a feedback control signal on a single input/single output loop.

In order to assure reliable flow signals, we chose the transition point from one-to-three element mode in a gap of 10-16% of steam range measurement of the steam flow, in this case between 50,000 to 80,000 pounds per hour (pph).

A critical operative situation appears when we reach full load and the feed water control valve is working near full opening; frequently, this introduces an oscillatory behavior in all control variables of master and slave controllers. This is because of the limited capacity of the feed water valve to control situations where disturbances need to respond beyond the valve capacity.

Another problem that easily appears is when natural frequency of the master and slave controllers is similar. Then oscillations behavior is present in all control variables.

New control strategy

We must program implantation in modern commercial control equipment. This gives good stability of the control system performance from start-up to full load, including the duct burner start-up and burner trip. It includes smooth manual-to-auto and auto-to-manual mode transfer. This is the same for one-to-three and threeto-one element transfer mode at any operating point. The feed water control system must live in harmony with the deareator level and pressure control loops. The control system must stand the start-up and shutdown of the other HRSG connected by a steam header.

To accomplish these objectives, we added to the conventional three-element control strategy a variable proportional gain as function of the level error in the master controller and a variable proportional gain as function of the steam flow in the slave controller.

ROME - An engineering contract that links a Romebased company that works with the CASE Center at Syracuse University with a number of other companies to provide spectrum-engineering support to a federal agency could eventually lead to industry-wide changes in the spectrum-management field.

ANDRO Computational Solutions, LLC is part of a team, headed by ITT Corporation, awarded a three-year, $147.9 million contract to provide engineering-systems support, technical analysis, test support, and long-term strategic planning for the Defense Information System Agency, Joint Spectrum Center (JSC) as it meets national security and military objectives related to the use of the electromagnetic spectrum. If all the options on the contract are exercised, it has a potential total value of $545.7 million. ANDRO’s president and founder, Andrew Drozd, hopes the contract results in a $1.5 million increase in sales for 2007, up from $1 million in 2006. By the end of three years, he would like to see an increase ranging from $3 million to $3.5 million.

Founded in 1994 ANDRO develops software used for electromagnetic compatibility. The software is used to determine the compatibility of all the various systems on a piece of equipment. For example, ANDRO does a lot of work for the military making sure a plane’s subsystems - such as antennas - don’t operate at the same frequency and interfere with each other.

n blackwater river estuaries, a large portion of external carbon, nitrogen, and phosphorus load are combined in complex organic molecules of varying recalcitrance. Determining their lability is essential to establishing the relationship between anthropogenic loads and eutrophication. A method is proposed in which organic C, N, and P are partitioned into labile and refractory forms, based upon first-order decay estimated by biochemical oxygen demand relative to total organic carbon, and C:N and C:P ratios as a function of organic carbon lability. The technique was applied in developing total maximum daily loads for the lower St. Johns, a blackwater Atlantic coastal plain river estuary in Northeast Florida. Point source organic nutrients were determined to be largely labile. Urban runoff was found to have the highest relative labile organic N and P content, followed by agricultural runoff. Natural forest and silviculture runoff were high in refractory organic N and P. Upstream labile C, N, and P loads were controlled by autochthonous production, with 34-50% of summer total labile carbon imported as algal biomass. Differentiation of labile and refractory organic forms suggests that while anthropogenic nutrient enrichment has tripled the total nitrogen load, it has resulted in a 6.7-fold increase in total labile nitrogen load.
Related Results: Commercial Modeling

* Global warming: she’s already
a modeling and TV sensation in…
* Software simulation tools for
EMC control of high-speed…
More

Video

* Rimmel Makeup RangeRimmel Makeup Range

More

(KEY TERMS: bioavailability; blackwater rivers; eutrophication; land use; nutrients; nitrogen and phosphorus loading; organic carbon; organic nutrients; source allocation; watershed development; TMDLs; water quality modeling.)
Advertisement

(ProQuest-CSA LLC: … denotes formulae omitted.)

INTRODUCTION

Accelerated eutrophication arising from nutrient enrichment of estuaries represents one of the most significant water quality problems within near coastal waters world-wide (National Research Council, 2000). Diagnostic and management approaches often rely on linking the sources, magnitude and timing of the external nutrient load to the response of the receiving water body with dynamic water quality process models. In the evaluations of external nutrient loads to estuaries in temperate climates, organic nutrients have typically not been differentiated with regard to lability, and organic carbon is rarely, if ever included (Jaworski et al., 1992; Magnien et al., 1992; Valiela et al., 1992; Boynton et al., 1995; Stepanauskas et al., 1999; Goolsby et al., 2001). For blackwater rivers of the southeast U.S. coastal plain, where a large portion of the nitrogen and phosphorus are contained within the organic fraction (Graves et al., 2004), and where natural organic carbon loads can confound the evaluation of oxygen deficits, the failure to account for differences in organic nutrient and carbon lability in the external load seriously compromises the assessment of anthropogenic eutrophication effects.

While inorganic nutrients and some low molecular weight organic compounds are readily assimilated by aquatic primary producers, organic nutrient forms, which must first undergo desorption, hydrolysis, bacterial decomposition or photo-decomposition (Bushaw et al., 1996) for inorganic nutrient regeneration and utilization, are less readily available. Organic nutrient re-mineralization is dependent upon the utilization preference of the parent substrate by general microbial heterotrophs (DeBusk et al., 2001). A general working framework has evolved that partitions organic carbon and nutrients into two pools: a labile pool, that can be utilized in time frames relevant to water quality processes of interest in the receiving water, and a refractory pool, that is decomposed very slowly and essentially inert for relevant time frames (Wetzel, 1990, p. 737). Carbohydrates, proteins, lipids, nucleic acids, and pigments, which are in higher proportion in younger, autochthonous plant material, typically represents the bulk of this labile pool, while humified, allochthonous OM, largely imported to streams as dissolved and highly colored, leeched, degraded terrestrial plant materials (colored dissolved organic matter, or CDOM), typically dominates the refractory pool (Meyer, 1990; Moran and Hodson, 1990; Kaplan and Newbold, 1995; Moran et al., 1999). In their work on piedmont and coastal plain blackwater rivers in the southeast U.S., Sun et al. (1997) demonstrated that the compositional changes that accompany diagenesis relate directly to lability, with blackwater stream CDOM appearing the most refractory per mole carbon, and this is in agreement with work that has shown some forms of soil humus in the allochthonous organic carbon pool to be decades to hundreds of years old (Raymond and Bauer, 2001). Although natural CDOM is generally believed to be resistant to microbial decomposition and largely unavailable for utilization by phytoplankton in typical estuarine residence times, these large, heterogeneous organic molecules contain a substantial amount of nitrogen (N), and to a lesser degree phosphorus (P), in their structures (DeBusk et al., 2001), and hence the sheer volume of this material with respect to other organic matter (OM) pools dictates that its relevance be considered.

Bob Restrepo lands at State Auto to rave reviews

This is the job I always dreamed of having,” says Bob Restrepo, who took over the helm of State Auto Insurance Cos. on February 10, 2006, making him the fourth Bob to head up the companies. He succeeded Bob Moone, who took over from Bob Bailey in 2000. Bob Pein was the founder of the company. But Restrepo, despite his surname, was an unusual choice in that he came from outside the State Auto fold. He serves as president, CEO, and chairman of the board of both the mutual company and its downstream stock company, State Auto Financial. Restrepo had spent much of his career at large stock companies, namely Aetna and Travelers where “I focused on personal and small commercial lines-businesses that depended on independent agents,” Bob says. “When I left Travelers, I went to Hanover. I like the regional company model. They have a unique focus on the market. When the boards offered me the opportunity at State Auto, I knew it was a dream come true.

“At their best, regional companies focus on commercial accounts that are beneath the radar of larger companies, as well as personal lines. These accounts represent a stable and predictable business that can be very profitable. When I looked at State Auto,” Bob continues, “I saw that it exemplified everything that was positive about a regional company. In addition, I saw some real opportunities to promote organic growth through enhanced pricing precision and new product development that relied on State Auto’s core strength-sound underwriting. This is a company that has boasted an underwriting profit in all but two of the last 15 years. And those two years showed an underwriting loss because of the merger with Meridian Mutual where we had to revalue some reserves. However, the core business was always producing combined ratios in the 90s or better.”
Advertisement

The business mix at State Auto is about 60% personal lines and 40% commercial lines. “I would like to see a 50/50 mix,” Bob says, noting that the company had been trying to get to that ratio for about 40 years. “We acquired several small regional companies that had personal lines books and this accounted for the bulk of our growth.” But that is something that Bob wants to change by achieving a better balance between organic growth and growth through M&A.

“You have to recognize that we’re in a relatively flat business environment so growth involves taking business from someone else,” Bob points out. “In order to accomplish that, we’re ramping up our investment in technology and new products.

“We’ve invested in a new personal auto product, CustomFit, that uses predictive modeling and will adapt the same techniques for other lines. The result is faster, more accurate pricing. We’re now extending that to include our BOP products and other commercial lines products where appropriate.”

While it has been less than a year since Bob Restrepo took office, agents already are enthusiastic about the new directions he is proposing.

“What’s happening at State Auto is all good,” says Bob Carmack, president and CEO of Carmack Insurance Group, Frederick, Maryland. “Our agency has represented State Auto for more than 50 years and always had a good relationship. Bob (Restrepo) is bringing fresh new ideas that will take the company to the next level. I’m really pleased with the efforts in the automation area. State Auto has always been a company that was easy to work with, but the new technology makes it even better. In addition to representing the company as an agent, I’m also a stockholder and I’m excited for that reason as well.” Carmack Insurance has premium volume of a little over $10 million, with a business mix of about 60% commercial lines and 40% personal lines.

“There’s a renewed spirit, new products, more competitive rates, enhanced automation tools, and broadened underwriting authority for the field representatives,” enthuses Craig Moon, president, Moon & Adrion Insurance Agency, Middletown, Ohio.

“We’ve represented State Auto for 85 years and we’re actually experiencing the excitement and enthusiasm of taking on a new carrier without any of the downside. Bob [Restrepo] is making a great company better. I have a renewed sense of excitement.

“Bob has met personally with the agents and with the people in the company. He’s spent a great deal of time finding out what they want and need,” Craig continues. “He takes action quickly where needed, delegates when appropriate, and has chosen to empower people so that they’re on board with the changes and also are working to make the company better. When I talk to him, I feel like I’m talking to a co-worker or buddy. He’s perfect for this job. The board hit a home run with this appointment. And the neat thing is that everybody feels this way, not just agents, but also the field reps, the people inside. I’ve even talked to people who don’t represent State Auto and they’ve heard good things.” Moon & Adrion is a State Auto Inner Circle agency, with a business mix of about 40% commercial lines and 60% personal lines.

Electronic speckle pattern interferometry (ESPI) was applied to noncontact, real-time evaluation of thermal deformation in a flip-chip solder joint. To measure the deformation of such tiny components as the solder balls in the flip-chip, the spatial resolution of ESPI was increased to submicron scale by magnifying the areas studied. Experimental-computational procedures were developed to obtain stress-strain curves for solder balls in the flip-chip based on finite-element modeling (FEM) of in-plane ESPI thermal displacement data. The stress-strain curve obtained for the flip-chip solder was compared with those for bulk solder. The microstructure was also studied to clarify the stress-strain curve results.

Key words: Flip-chip, electronic speckle pattern interferometry (ESPI), finite-element modeling (FEM), stress-strain curve, bulk solder, reliability
Advertisement

INTRODUCTION

In the quest for enhanced performance and function in electronic devices, electrical component size has been decreasing, and the number of inputs/outputs increasing. To accommodate these changes in electronic packaging, solder-joint sizes have decreased, and the joint numbers per package have increased. Such trends in electronic packaging have raised concerns about solder-joint reliability, particularly the fatigue failure of solder joints induced by temperature cycling.

Various experimental reliability tests are performed to predict the long-term reliability of a solder joint.1,2 Because such experimental work is expensive and time consuming, finite-element modeling (FEM) analysis is increasingly used as an effective alternative.3 Such finite-element analyses must have accurate material properties as input if the results are to be accurate. However, unified material properties for the solder joint are difficult to determine for a number of reasons: the difference between micro and bulk solder-material properties,4 the strong dependence of solder-material properties on manufacturing and aging conditions,5,6 and the lack of testing equipment and standards for microscale solder materials.4,7 Somewhat surprisingly, studies have used different material properties in modeling even the same solder material.

The objective of this study is to obtain stress-strain curves for Sn-36Pb-2Ag solder balls in a flip-chip using FEM of thermal displacement data, as measured by electronic speckle pattern interferometry (ESPI). The stress-strain curve obtained for the flip-chip solder is then compared with those for bulk solder.

MEASURING THERMAL DEFORMATION OF THE FLIP-CHIP BY ELECTRONIC SPECKLE PATTERN INTERFEROMETRY

In-plane thermal displacements for flip-chip solder joints were measured by ESPI during heating from 25°C to 125°C. These values will be used to determine stress-strain curves for solder balls in a flip-chip through FEM.

Electronic Speckle Pattern Interferometry Measurement Technique

Noncontact optical methods are desirable in measuring thermomechanical deformations because they allow real-time, whole full-field measurement during operation. Different but complementary optical methods include holographic interferometry,8,9 moire interferometry,8,10,11 and ESPI.8,12 Of these methods, ESPI is among the most promising for measuring thermal deformation in a flip-chip package in that it requires little or no special specimen preparation and can measure in-plane and out-of-plane deformation with high sensitivity.

Figure 1 is a schematic of the ESPI system for measuring in-plane deformation. The ESPI measurement is based on the interference of two speckle patterns recorded before and after deformation. The speckle patterns arise by the interference of two incident beams, an object beam and reference beam (one of the two object beams in Fig. 1 functions as a reference beam). As the specimen deforms, the resulting surface deformation changes the phase difference between the object and reference beam and thus alters the speckle pattern. Subtracting the deformed speckle pattern from the undeformed produces correlation fringes that yield a displacement field through the well-known relationship between fringe order and displacement.8

An ESPI system was applied to measure in-plane thermal displacements in a flip-chip, which are small and submicron deformations. To improve the spatial resolution, a long working-distance microscope (focusing distance: 34 mm, magnification: ×10) and a zoom lens (maximum magnification: ×2) were attached to a charge-couple device camera. By adjusting the two lenses, images 20 times larger than previously obtainable were generated, and the optimal focus distance and light intensity could be controlled. In addition, a phase-shift technique was used to improve the precision, convenience, and usefulness of the interferometry.8 Most measurements were made with the aid of an image-processing computer.

Flip-Chip Specimens and Thermal Chamber

For the flip-chip specimens, a Si chip was attached face down on the substrate (FR-4) by reflowing solder balls on metallized terminals on the chip and substrate. The under-bump metallurgy (UBM) was constructed by sequentially plating copper (2 µm), nickel (3 µm), and gold (0.1 µm) on both the chip and substrate.13 The Sn-36Pb-2Ag, 500-µm-diameter solder balls were used. Solder reflow was performed at a peak temperature of 240µC for dwell times of 80 sec. To measure cross-sectional deformation, a cross section was created through the outermost row of solder balls and polished with 200-grit paper.

Remember when only a few banks and financial companies could originate major syndicated loans and asset-backed financing packages? Today there are dozens, and the list is growing fast. Competition is intense; margins have been shaved to the bone. And that was before Enron.

In this business environment, it’s obvious that decision makers at every stage in the lending process need timely, actionable business intelligence-the kind of up-to-the-minute decision support that I characterize as immediate information. But where can commercial bankers go to get it?

One good place to begin is with their own lending systems. With foresight and the right tools, commercial lenders can leverage their loan-processing installations into decision-support systems that deliver detailed, credible, up-to-date information about profitability, portfolio risk, and exposure. Here are the basic tools and principles you’ll need.

A Separate, Historical Database and Server
Advertisement

Most lending systems store data for only a few periods. Some of your customers have been with you for years, and you want to keep them for many more. Data storage is cheap and historical information can be valuable. It may, for example, reveal how a profitable customer relationship is becoming unprofitable.

Business intelligence tasks like modeling and forecasting place enormous loads on a database. So it makes sense to keep the historical database separate from the operational loan-processing database. It also makes sense to keep it on a separate server, or if you’re processing on a mid-range or mainframe, in a separate partition. Separation can also eliminate contentions for system resources and performance degradation on your production databases.

End-of-Day Extracts and Interfaces

Why would you want to start analyzing today’s problem with last month’s data? You’re already running the EOD routine anyway; why waste fresh data?

Preestablished, Multidimensional Data Hierarchies and Business Models

The basic structures for most forms of business analysis are common from task to task and bank to bank. If you’re doing a profitability analysis, at some point you are going to roll revenues up in one place and costs up in another, and then subtract. Whether you’re looking at customer relationships or products or business units, the basic methodology will be the same.

It’s expensive and time consuming to reinvent the wheel for every analysis. Using pre-adapted business models will save time and money. If you use multi-dimensional structures and models, then you will be able to rotate, pivot, drill down (”slice and dice the information”), or aggregate according to the demands of the task.

Off-the-Shelf, Multidimensional Data Analysis/Presentation Tools

There are at least a half dozen strong OLAP/MOLAP solutions out there from providers like Oracle, Business Solutions, Cognos, or Brio. They feature intuitive GUI front-ends and user-defined options. And they’re all a lot cheaper than anything you could build in-house. Pick one and get to work. In fact, you could even pick two: one for the power users who do the number crunching; another for executives who want to review results fast to see what they have to take care of today.

Choose Web-enabled Solutions

Information is genuinely valuable only when it is in the hands of the people who need it to make decisions and take action. Make sure that report writers can generate reports in HTML so that they are quickly distributed to users in distant offices or easily read by traveling executives.

Aim for Self-Sufficiency

Few business analysts can write queries in SQL, and I doubt that any commercial banking executives can. You need tools that let users try different approaches. If you have to wait weeks for the IT department to provide your data, you’re dead meat for more nimble competitors. Immediate information means information you get and use now.

It makes good business sense to get more value out of the data that your enterprise is spending so much money to process. With ingenuity and good tools, today’s advanced commercial banking systems do a lot more than process loans. And one of the things they can do is provide the immediate information your business analysts and executives need to thrive in a dangerous, volatile economy.

There are significant differences between analytical modeling and simulation modeling, particularly in the areas of scalability and complexity.

Analytic modeling, a mathematical representation of computer or other systems for performance analysis, has limited ability to accurately model complex or dynamic environments. Single hosts or nodes in a simple platform can be quickly modeled through mathematical means, by having a large number of users generate an “average” size transaction against a single server. Queuing theory equations and formulas are used to estimate the performance or response time of an existing or planned system by relating input parameters to obtain output statistics.

While these types of models can be created quickly, they are unable to accurately handle the following types of systems.

Concurrent access to internal resources (memory, disk, etc.)
Advertisement

Prioritization of traffic streams or processes.

Background or less significant communication protocols between components.

Interrupts and the ability for one process to block or impede another.

Complex events which may shift or vary over a known time interval.

Validation of inter-dependent functions or processes.

Systems with variable loads, spikes, unpredictable client interaction.

These limitations force the use of approximation and estimation techniques, increasing the likelihood of inaccurate results. In addition, analytical models cannot be effectively used for the following types of project work:

Where a new design is under consideration, and no baseline model exists;

When a component upgrade or change is under review and the effect on performance needs to be determined;

When many interdependent transactions exists within a system and the effect on response time when each are competing needs to be measured.

In an analytic model, all traffic is represented at a single level and is very effective for simple client/server or mainframe environments where load and traffic volumes remain fairly consistent. These narrowly focused and simple models are well suited to existing environments which remain static with little variation.

The Simulation Side

Simulation modeling, in contrast, offers the ability to create a valid representation of an entire complex system consisting of all computers, routers, bridges, applications, and database servers. In the simulation, the interaction of all these components can be launched and the results measured under various scenarios. Simulation modeling can be used for any system, but is particularly well suited for complex environments. In detail, it can be used to do the following.

Maintain accuracy by scaling and reacting to system or platform variations.

Provide valid results for all types of software applications, both in- house and commercial.

Address capacity planning questions at both the hardware and software resource levels, while allcwing the connection between the two to influence the results.

Represent workload fluctuations on a minute by minute basis, with batch arrivals and processing placed in parallel with interactive traffic flow.

Create accurate models of truly concurrent systems when; multiple threads may exist within one or several components.

Represent database locking and I/O processing with greater accuracy.

Provide a limitless approach to modeling any system regardless of complexity.

Since the majority of today’s computer networks are dynamic in design, one small change in a single component or process can trigger a much larger change and have a negative effect on end user response time and system availability. These small errors or miscalculations may be compounded through the planning and design process, until the cost of correction within the finished system becomes excessive. Simulation modeling can highlight these errors and eliminate them before the system or platform is finalized.

Different Testing

Another method of performance analysis is through the use of load or stress testing, where system behavior is observed and recorded under predicted user volume. This type of testing usually involves the use of tools which can create groups of virtual users or agents which mimic real clients running actual application tasks.

For example, if a product manager needs to know if 100 users can update a Web server inventory database at the same time, he would request a load test matching those requirements be run against the server and application.

During the test, server and client resources, response time, etc. would be recorded for analysis.

While load or stress testing can be helpful, the process does have several limitations.

* The system or platform under review must have already been designed and implemented into production.

Choosing products to launch from a set of platform based variants and determining their prices and launch sequences are some of the critical decisions involved in any new product development (NPD) process. In this paper we present mathematical models which facilitate such decision making. The products considered are commercial vehicles, and representative data from a commercial vehicle manufacturer in India have been used for analysis. Our model which determines launch sequence as well as price of the products provides useful insights on the impact of economic conditions like boom or recession on prices, and also cannibalization.

1 INTRODUCTION

In a competitive market firms have to cater to the needs of the customers by offering a variety of products, and at the same time keep the development and manufacturing costs low. Many of the successful firms, particularly in automobile and consumer durable industries, owe their success to developing an effective platform from which they are able to launch a series of derivative products. But the firms need to judiciously balance the conflicting marketing requirements of variety on one hand and complexity of operations and development, and hence, escalating costs on the other. The common trend in automobile industry is to organize design groups on the basis of product platforms. In many of the firms, including the one under study, Vehicle India Ltd. (name disguised for confidentiality), a separate group consisting of personnel drawn from marketing, design, engineering and process planning is responsible for the critical decisions in any development program. We consider the intertwined decisions involved in NPD projects starting from choosing a set of platform-based variants to determining their launch sequence and prices. These decisions are taken after the features of product concepts have been finalized based on market responses and tools like Quality Function Deployment (QFD).

We consider the platform extension problem from the point of view of decision makers evaluating NPD projects, and add to the literature by considering not only development costs but also plant configuration change costs, productionising and technology acquisition costs, complexity costs of maintaining the variants, and labor and material costs in a mixed integer linear programming (MILP) model. The model, instead of evaluating each product development program individually, helps in allocating manpower for each department and choosing the products together subject to the resource and project duration constraints. For the platform extension problem, our endeavor is to determine the platform extensions to be launched for each platform, given that the base models will always be launched. The objective of this simple model is to provide the managers with insights on the binding constraints for each variant and the threshold level of cannibalization that can be allowed. Evaluating new products individually on the basis of net present value as done by Vehicle India Ltd., does not take into account the sharing of resources by the products, and run under the risk of providing a suboptimal solution. The comprehensive MILP model of ours, which is solved using GAMS 21.0, helps in choosing the optimal set of products, considering the relevant costs and satisfying the constraints. Given the available manpower in design and technical support departments, the firm has to satisfy the manhour requirements for design, prototyping, testing and production ramp-up to launch the products on time. Our model helps managers take decisions about the manpower requirements for different departments.
Advertisement

Appropriately pricing the platform based variants and determining their launch sequences are another set of critical decisions faced by the firm. Sometimes the launch dates of new products may be externally enforced as in the case of launch of vehicles satisfying latest emission norms. But the firm can exercise the options of staggering the launch of various products and price them appropriately according to the boom and recessionary conditions in the industry at the time of launch. Several researchers from operations, marketing and economics have studied the problem of determining price and launch sequence. Seminal work [9] helped in choosing between simultaneous and sequential product launches depending on the varying levels of cannibalization. In [7] they made a significant contribution to this class of problem by developing launching sequences for platform-based products. The impact of concurrent technological development on launch sequences for high technology products was studied in [1]. But the impact of economic conditions on launching sequences and prices does not seem to have been considered in the literature before. The motivation of the study stems from the following questions

Reliable loss prediction requires both robust estimation methods and accurate data. This article presents a way to leverage ratings agency data that can provide greater flexibility and stability of results in simulation-based estimates of future portfolio losses.

Based on a simple behavioral model that quantifies the structural relationships in historical default frequencies and transition rates for different ratings, (1) this technique leads analysts to hypothetical transition matrices for portfolio loss simulations that preserve the basic relationships observed in the historical transition and default rates reported by the ratings agencies, allowing for unlimited sampling. The matrices can also be linked to macroeconomic factors to mimic the dynamics of credit cycles and economic shocks, allowing for richer descriptions of plausible future scenarios and what-if scenario analysis that goes beyond the limitations of historical data.

The Basel II capital adequacy framework provides strong incentive for financial institutions to use internal risk management systems to measure risk and determine sufficient regulatory and economic risk capital. While commercial risk measurement tools can be used as part of an overall solution, institutions must tailor them to their own portfolio specifications. Further, some of the development and implementation of the new systems will fall to their own risk management teams.
Advertisement

In many cases, whether they use commercial models or internal methodologies, analysts continue to rely on data from the major ratings agencies for default rates, ratings migration rates, and other key statistics. Despite recurring and somewhat troubling issues regarding the meaning and consistency of ratings, regulators tend to be more accepting of methodologies based on agency data because of the agencies’ long and well-documented ratings histories. This data may indeed be deeper and may conform better to an accepted standard than banks’ own internal ratings histories, yet the depth of agency data generally falls short of what’s needed for the Monte Carlo-based economic risk capital estimation techniques in widespread use today.

The Shortcomings

The simplest portfolio loss model assumes that ratings transition probabilities are stable across obligor types and across the business cycle, and that a single set of average historical ratings transition and default rates is all that’s needed to characterize potential future losses. However, there is ample evidence that credit migration and the ratings process depend on a number of factors, such as the state of the economy-for example, the probability of downgrades and defaults is greater in a downturn than in an upturn. Moreover, historical data is volatile; thus, the average-rate approach will understate potential tail loss–the very thing we want to measure with precision. A slightly more sophisticated alternative is to use observed annual historical-rating transition rates as a sample from which to draw plausible future credit migration scenarios to simulate the forward loss distribution. The main drawback of this method is the small number of historical-rating scenarios on which to draw. Accurate Monte Carlo simulations for large portfolios usually require tens–or even up to hundreds of thousands–of random draws. However, because historical scenarios number only in the tens, the simulated loss distribution will tend to be lumpy as tail losses bunch up around the worst year from the historical period. Clearly, this problem cannot be overcome by increasing the number of Monte Carlo simulations.

A Behavioral Model of Risk Perception

A different approach is to directly model the relationship between transition probabilities and macroeconomic factors and then simulate plausible ratings migration patterns over time by generating various macroeconomic conditions. To do this, we need a behavioral model of how risk ratings are assigned. Let’s begin with the observation that ratings are opinions of credit quality, representing different degrees of belief in the credit quality of the firm. Agency statistics, such as default and transition frequencies, are merely by-products of this rating assignment process, rather than properties inherent to the ratings themselves2 Analysts’ judgments, meanwhile, are based on a combination of qualitative and quantitative comparisons of the credit risk they perceive. Even if specifically attempting to arrive at a default-probability calculation, the analyst cannot be sure of the precise relationship between the risk factors affecting the obligor and his or her own mental model of risk perception, which may lead to errors in risk assessment. Thus, even with complete and perfect information on the obligor’s risk exposure, the analyst would still face “model risk” because of judgment. Any qualitative comparison between two risk exposures is clearly probabilistic in nature since it relates to uncertain events. Unfortunately, analysts’ perceptions of the probability of default, expected losses, and future ratings revisions are not publicly available and therefore cannot be tested. However, we can construct a behavioral model for the average perceived risk that can be calibrated with historical default and transition rates associated with a given risk perception (rating at a given point in time) assuming that the ratings are unbiased estimates of the average (ex-ante) analyst’s perception of the risk criterion