MapleSim LabVIEW/VeriStand Connector

The MapleSim LabVIEW/VeriStand Connector lets designers bring models built using Maple's symbolic mathematics engine into the LabView environment for more rapid prototype development. Image: Maplesoft

Running a lab lean, fast, and green is about more than choosing the right personnel, facilities, and equipment. Increasingly, software can make the key difference.


The virtual working environment—a fancy term for the computer desktop—can be a sensitive subject for designers and engineers. It is the place where most of today’s developers spend much of their time, and it is rapidly becoming the place where valuable time and effort is lost in the research and development process.

Often a design problem is so complex—designing a new molecular structure, or building a hybrid engine system—that unless the correct software solutions are chosen, years of effort can be lost. Finding the best solution isn’t easy, but vendors have been ramping up their efforts quickly to become part of a successful laboratory workflow instead of becoming a roadblock.

Building the product: From virtual to reality
Prototyping is a common difficulty for labs engaged in product development. While designing a part or a mechanical system can be carefully done with the assistance of CAD/CAE tools and/or advanced analytics and simulations, ensuring that all of a complex devices systems work together can often lead to wasted time in broken prototypes, reverse engineering, and overuse of materials.

One leading effort to streamline this process is LabView, from National Instruments, Austin, Tex., which has grown to take advantage of the proliferation of embedded systems. Kurt Williams, product manager, LabVIEW Industrial, says that LabView was originally conceived to take advantage of the surge in embedded systems. Nobody really knew how significantly a graphical system design language would impact productivity until real-world applications were pioneered.

Illustrating LabView’s flexibility are two highly different projects using the graphical design interface. One project involved quickly developing and deploying an embedded, multimodality diagnostic imaging system for small animal veterinary practices. The development work began with LabView and CompactRIO processing equipment to create a functional prototype. Once the optimal system design was reached, the entire prototype software code was quickly ported to a different NI logic platform to finalize the solution for the company, Animage LLC. According to Williams, the approach saved about three years of development time, amounting to hundreds of thousands of dollars.

LabView also supported National Instruments’ development of wireless sensor network (WSNs), which were designed to address a broad range of remote and environmental monitoring applications for its customers. As part of their shakedown process, NI monitored a pond at its headquarters in Austin. Before a WSN system was available, installing a wired system to automate measurements at the pond would have required running wires under the entrance to NI headquarters, disrupting thousands of employees and costing too much money. The ability to connect to measurements wirelessly enabled the team to avoid running power and communication cables. The instrument design project, which did include NI CompactRIO processing, was performed with LabView software.

“People are always looking for ways to get to market faster. Over time, people are beginning to see the value of LabView and the ability to prototype stuff on LabView and get real-time performance,” says Williams. Once a developer gets to know the LabView environment, the ability to program visually maximizes the ability to make both analytical and intuitive changes.


Embedded system testing

Embedded system testing, such as this vehicle stability program using MapleSim, can speed lab work. Image: Maplesoft

There are other important factors that help LabView’s case, says Williams, including the need to move an idea to market in the interest of protecting the intellectual property that enables it. “There is also always a desire to prototype and prove a concept as soon as possible, especially in the medical field.”


In that vein, efficient development will likely encourage the use of modeling in close conjunction with virtual design environments. After witnessing the diverse uses for LabView—solar-powered cars, manufacturing robotics, and wind-turbine control systems—National Instruments has strengthened its product’s ability to work with popular modeling solutions.

“I definitely think simulation will continue to grow. Developers will test out control algorithms on something that is simulated rather than a real-world device. If there are mistakes in the design process that would damage a physical prototype, they will be caught in the modeling stage,” says Williams. That will in turn save costs in addition to time.

Part of the move toward virtual, simulated environments is the demand for accuracy in mature industries. In vehicle development, for example, fuel management for combustion engines requires exceptional precision several orders of magnitude higher than even 10 years ago. The ability to model these phenomena are now possible, but integrating such models in a workflow that results in a prototype hasn’t been straightforward.

Enter Maplesoft, Waterloo, Ont., which introduced MapleSim late last year as simulation support for the well-established Maple symbolic mathematics platform. The potential impact of a symbolically-driven simulations tool on the customer was anticipated, but what the vendor did not expect was the rapid way even larger industries would benefit from its capabilities.

“What was a really wonderful surprise for us was the ability to use symbolic mathematics to create models. You can actually simplify the calculations,” says Tom Lee, vice president of the Applications Engineering Group at Maplesoft. The reduction in computing effort resulted in real-time simulations that would run about 10 times faster than when using conventional algorithms. “Without that speed-up, we couldn’t do some hardware-in-the-loop systems, which of course help when it comes time to actually start building a prototype.”

A good example of how simulation aids the engineering process is static forces versus dynamic forces. Direct mathematics can efficiently identify optimal specifications for a given part, such as a gear, but the dynamic characteristics are often too complex to justify rigorous testing. Estimation and wasteful overengineering are often the result. In 2007, Toyota began using Maple’s symbolic engine to refine its engine technology, making great strides in prototyping speed and efficiency, reports Lee.

“In a nutshell, the way we approach it mathematically makes certain simulations feasible. Now the symbolic math is becoming fully mainstream, and there are really creative ways to apply that math,” says Lee.

Lee says that long before MapleSim was launched the concept of integration with LabView had been discussed. As a result, the MapleSim LabVIEW/VeriStand Connector was developed to allow the export of MapleSim models to The MathWorks’ (Natick, Mass.) Simulink formats, which are compatible with LabView. This includes rotational, translational, and multibody mechanical systems, thermal models, and electric circuits. These symbolically-derived models, using what are called S-Functions, retained the ten-fold speed advantages.

Similarly, The Mathworks simulation tools have been saving time for engineers in a variety of disciplines. Engineers at Argonne National Laboratory, Argonne, Ill., designing next-generation fuel cell, hybrid electric, and plug-in hybrid electric vehicles, have used Simulink and other MathWorks software tools to build a Powertrain System Analysis Toolkit that helped them optimize fuel economy and performance while containing development costs. They were able to simulate multiple powertrain configurations without building costly physical prototypes. With distributed computing, they were able to run about 3,000 vehicle simulations in less than 48 hours. With one computer it would have taken them up to five months.

“This approach was able to give us much more accurate information. It was a very important advance. Now for one technology we can provide much more data associated with it because we are running so many more simulation,” says Sylvain Pagerit, manager of vehicle development at Argonne.

Lab operations elevate skyward
Another efficiency measure labs can adopt with regard to software is cloud computing. This colorful term has become the buzzword for the rapid and inevitable move toward a centralized data processing infrastructure. How quickly is it happening? According to recent market reports, the cloud computing sector has continued its growth throughout the recession and is accelerating now that hosting projects that were put on hold during the recession are now coming to fruition. The IT analysis firm IDC predicts a 4.5% growth rate for “hosting” in 2010. The business is already into the billions of dollars and so-called Software-as-a-Service (SaaS)—that is, software that is accessed and used via an Internet connection—is proliferating.

Why is it so attractive? Is it easier for a small lab to purchase a few dozen desktops, a server, and software that needs frequent updates, or is it easier to purchase one or two desktops, several notebook computers and one or two integrated solutions accessible via the Web?

More and more laboratories are adopting the latter, and companies that cater to specific cloud computing needs are popping up everywhere to meet this demand. The biotechnology and pharmaceutical labs were among the first to make this transition, and as additional technologies such as desktop virtualization and thin client accessibility grow, labs may one day be forced to change. One such company is Sciformatix, Los Gatos, Calif., which has bridged security and bandwidth concerns and put real-time lab management entirely in cyberspace with its first product, SciLIMS.

Tom Kent, president and CEO, equates his company to the top level of Maslow’s hierarchy of needs pyramid, in this case applied to a biotechnology laboratory. The basic methods in the lab represent the bottom level; the inventory management of consumables or compounds is the second level; the LIMS (laboratory information management system), with instrument integration, is the third level; and the SaaS solution, he says, is at the top.

“Whether a lab thinks about it as such, every lab needs to maintain information tracking on subject in the lab. Ultimately all of them need that bottom layer or two,” says Kent, but in terms of formal LIMS activity, many labs are still stuck with informal tools such as Microsoft Excel spreadsheets.

There’s nothing wrong with these tools themselves, says Kent, but the difficulty of sharing and using them in a fast-moving lab can mount quickly.

“About 80 to 90% of labs operate on that basis without a formal LIMS solution,” says Kent. Many of them, he believes, could benefit from key advantages offered by a solution like that offered by Sciformatix: rapid startup; ubiquitous access; absence of installed, purchased, or maintained computing infrastructure; and cost reduction. The cost savings is one of the most powerful arguments for SaaS: an organization that moves to SciLIMS should expect at least a double-digit ROI on an annual basis and sometimes a triple digit return, according to Kent.

He answers the inevitable question of security in this way: customer-relation management solutions have now moved to hosted platforms, and these often represent the most sensitive kinds of data.

In addition to supplying real-time access to any portal with an Internet connection, a software service such as SciLIMS is attractive because of its search capabilities. For example, finding the melting temperature of a compound is as simple as searching for certain criteria and being supplied with the resulting information. It can be integrated with barcode information to help locate the physical sample if needed, and analytics are available to generate custom reports.

“We’ve built the ability to literally tailor SciLIMS. Users can determine and set up the sample and specimen and they can also define the specific properties to capture and track with those sample types,” says Kent, who, in the two months since SciLIMS was officially launched, has already helped at least one lab abandon an antiquated hand-written labeling scheme.

Sciformatix has designed its product to deliver a quick solution to a start-up or small to medium lab operation looking for convenience, customization, and ease-of-use. Some larger labs may require a powerful solution that can accommodate massively parallel experimentation and manage it through common programs such as Excel and email. Others already use Waters’ Empower or Agilent Technologies’ Open Lab and wish to access that information from an electronic laboratory notebook. Additionally, transparency is accepted to be a primary driver of scientific discovery, and tools that help share information, within the company or with other companies, will aid discovery and development.

“If you are doing 100 or 1,000 experiments per day you must have informatics in place to manage the data and the process,” says John McCarthy, vice president of product management & strategy at Symyx, Sunnyvale, Calif. “Without informatics in place you simply drown in data and spend 100 days preparing and analyzing your experiment that you took one day to run.”

Productivity needs informatics, says McCarthy, which is why his company has spent considerable effort not only developing the equipment and tools that let lab professionals execute a process, but also building an online tool chain that expedites the analysis and utilization of the data these experiments produce.

Primarily directed at life science and chemistry labs, Symyx’s products are augmented by its solutions in cheminformatics, laboratory operation informatics (LOI), and ELN.

“While Symyx Notebook receives much of the publicity because of the excitement around ELNs and the recognized 20% productivity improvements in moving from paper to electronic environments, Symyx and our customers see our Laboratory Operations solutions as a mechanism to further improve laboratory efficiency,” says McCarthy.

Drug discovery laboratories have long been characterized by painstaking research that relies heavily on repetitive processes such as assays. Attempts to streamline the process have ushered a variety of creative solutions, and software is one of them. Typically, Symyx’s LOI products are used to assist in chemical registration reagent inventory management, sample management, and, most recently, work request management and equipment and metrology solutions.

“We think of it as bringing supply chain approaches to the drug discovery process,” says McCarthy.

Symyx is no stranger to hosted environments, either. Whereas Sciformatix’s customer are typically smaller labs looking for an inexpensive and often first-time SaaS solution, Symyx must consider the sometimes substantial data needs of its customers when delivering its software in a hosted environment. This includes hosting its customers’ chemical registration data.

“While many organizations are flocking to a hosted model to lower their initial capital costs in hardware, software or FTE costs in managing and maintaining the environment, we see this as just the initial benefit. The true benefit to scientists is enabling them to collaborate more efficiently,” says McCarthy.

Collaboration, he continues, is more important with the modern research laboratory where a lab across the world may synthesize the material, another lab characterize the material and a third location measure bioactivity. A hosted solution lets that process become more efficient.

Symyx has also found it desirable to begin its own collaborations with industry players such as Thermo Scientific. This was the smart direction to take, according to McCarthy, because it allowed Symyx’s Notebook product to fold into Thermo’s customers software solutions with disrupting the day-to-day activities that involve Thermo’s own Watson LIMS products. Secondly, Symyx wants to make Watson LIMS data more readily available to principle investigators. In effect, Notebook, says McCarthy, is able to put usable data into decision makers hands more quickly.

“This is the first of several points where Symyx and Thermo Scientific are integrating our products. We will provide integration between Thermo Scientific’s Atlas product and Symyx Notebook in January 2010 to support analytical chemists. And we are looking at further integration with Thermo Scientific’s Sample Manager, Galileo, Grams and others,” says McCarthy.

Kent believes the larger LIMS companies, like Thermo Scientific, will have a harder time moving to a SaaS-only platform.

“The biggest challenge that any LIMS vendor will face in terms of moving to SaaS is to provide infrastructure of software that will allow simple tailoring software,” he says. Simply put, size may force them to adopt a “partly cloudy” approach to hosted products.

The need for collaboration, however, is deeper than simply pairing products together and ensuring that lab software, SaaS or otherwise, is easy to use and is compatible with common file formats. There is a pressing need for a modern lab environment in which software, hardware, techniques, and know-how are integrated. It is the reason why the adoption of portals and collaboration tools such as Microsoft’s SharePoint is accelerating and why hosted solutions are finding customers.

“The modern electronic laboratory environment is being driven by scientists’ desire to spend time doing science rather than moving files around or copying data in those files,” says McCarthy.

Published in R & D magazine: Vol. 51, No. 7, December, 2009, pp.12-14.