Originally Published IVD Technology October 2004
Electronic and mechanical components and software
 |
| Rigid-flex circuit geometries by HEI Inc. (Boulder, CO) enable smaller packaging for implantable, handheld, and point-of-care diagnostic devices. |
A company approaching the development of an IVD instrument should make sure
that supply-chain management is actively undertaken from early in the process.
Marketing managers often point out to the engineering staff that time to market
is critical to success. Because of that, a sound supply-chain management plan
is vital to ensuring that a device can be introduced with no significant loss
of time.
Supply-chain management at any company occupies a link between the engineers
designing a new system and the vendors of the sourced components. Managers will
find that a close working relationship with both vendors and the engineering
staff optimizes supply chain processes and pays great dividends during the transition
from development to manufacturing. Organizations are constantly looking to improve
margins through lean-flow manufacturing, kanban management, and product synchronization,
not to mention the selection of cost-effective components. These techniques
often involve maintaining good vendor relationships.
The engineer’s perspective is different, and involves a different challenge.
The technologies associated with microfluidics, DNA/RNA, microarrays, and fluorescence
in situ hybridization (FISH) drive critical part selection during the product
design stage. Component selection and the system integration associated with
that component are the engineer’s chief concerns.
System integration involves all engineering disciplines, including the validation
and verification effort and the manufacturing process. All too often, designers
lose sight of these two project areas. More important, V&V and manufacturing
are not even acknowledged in early discussions with vendors during establishment
of the supply-chain management plan.
Technology Trends
To those taking part in the effort to isolate the human genome, this may be
the postgenomics era. On the other hand, those attempting to utilize this information
for clinical purposes might see this as the beginning of the genomics age. It
is clear in any case that the identification of genes, including their expression
and resulting proteins, is going to be a large part of the future for in vitro
diagnostic medicine.
Significant development efforts are under way in this area, involving microarrays,
FISH, cytology, and various forms of mass spectroscopy. These are largely focused
on the biotechnology and drug discovery markets, but there is a strong underlying
motivation to bring these technologies to the IVD marketplace as quickly as
possible.
Another area of technology focus is microfluidics and biochips. The concentration
in nucleic acid (DNA/RNA) has been a cause of concern for the reagent side of
the business because the reagents used in that field are more expensive than
traditional chemistry and IVD reagents.
A third area of heightened activity in the IVD market, although somewhat underreported,
is associated with homeland security and bioterrorism. Significant government
funding has occurred in the past year to support this effort. Companies working
on these technologies use all of the scientific techniques named above, and
more, in developing systems it is hoped will be able to identify biological
agents quickly, with great sensitivity and specificity. It has been said that
what NASA did for the semiconductor industry, biodefense will do for the diagnostic
market.
The upshot of these trends is that “smaller and faster” are themes
that will be prominent in the future. Integration of mechanical, electronic,
and software components in creating devices to address the coming technology
challenges is crucial. Microarrays will require finer-resolution optics, as
will FISH and cytology. All of these technologies militate toward a microfluidics-based
technology to reduce cost and system complexity.
The Integration Challenge
Most instrument development projects begin with the establishment of the general
system architecture. Here, the engineering staff divides the system into subsystems
that address specific areas of functionality. During this process, the system
engineer considers what general hardware, software, optics, and electronics
can address a particular need. Additional considerations are serviceability
once the system reaches production. This architectural process is very fluid
and dynamic. When the general structural plan has been settled, the engineers
begin to look at alternative hardware solutions and to balance their pros and
cons.
Orders of components used in forward-looking diagnostic instrumentation increasingly
are placed with vendors that are able to customize their offerings. Customization
requires frequent and timely communications between manufacturers and vendors.
The challenge for the instrument developer is to select the component that minimizes
the integration demand.
There are trade-offs that need to be considered. For example, when implementing
a stepper motor driver, a motor controller could either be purchased from a
different company and integrated or be developed internally. Alternatively,
a stepper motor with a built-in motor controller could be purchased in order
to reduce project risk and accelerate the schedule. However, this would entail
costs 15–50% higher. In another case, developers might identify an input/output
(I/O) board that includes features similar to those needed for the instrument
and compare this option with designing a custom I/O board to interface with
peripheral devices within the instrument. The cost of the former option may
be higher because the board has features not required for the application and
the margin to the organization supplying the board. The lower-cost latter option
will require more development time but provide a more custom and application-specific
solution. In considering choices like these, sales projections for the instrument
should be measured against the costs of in-house development versus purchasing
an off-the-shelf solution to understand the most cost-effective approach.
As mentioned, microfluidics technology is now central to many diagnostic instruments
as the industry matures and miniaturizes. The capability of an instrument to
move fluid in nanoliter volumes grows increasingly desirable. Subsequently,
focus is directed toward pumps, improved precision and accuracy, and more-sensitive
liquid-level-sensing circuitry. The technology for dispensing microliter to
nanoliter volumes requires different components than the traditional syringe
and stepper motor. The demands on motor controllers and encoders are much greater.
The integration challenge here is the mix of technology—the right hardware
and the right electronics with the right software—that will ensure that
project requirements are fulfilled in a timely manner for all involved, including
the validation and verification team.
Incorporating Other Technologies
 |
| A sample carrier carousel by HEI Inc. (Boulder, CO). Automating laboratory assays requires excellent communication of process and scheduling requirements. |
In order to develop a fully functional system including transport mechanisms
for samples and cuvettes, temperature regulation for assay reactions, reagent
cooling, and, depending on the technology involved, even thermal cycling to
80°–95°C, many other subsystems besides microfluidics require some
form of active control. Improvements in flex circuitry and thick-film technology
have provided advances in this area in recent years—advances that address
both price points and the design of the heaters, coolers, and temperature-monitoring
technology employed in applications like thermal cyclers for DNA/RNA.
Industry miniaturization efforts have been accompanied by a need for greater
motion control precision in applications like microarrays and FISH. Thus, encoders
with nanometer resolution are now part of many systems, especially those that
require imaging in pixels as small as 0.01–1 µm. The introduction
of such encoders has increased processing demands and information management
overhead, which requires more processing power at the controller level. Whenever
there is significant motion control involving an encoder with nanometer precision,
the best solution, from the standpoint of cost and the development schedule,
is to purchase a dedicated controller as one of the system components.
Not to be overlooked are larger subsystems that integrate components into subassemblies.
Examples of these are fluidic pipettors, microtiter plate stackers, and electro-optical
imaging subassemblies. Again, the balance between cost and schedule should be
considered. While these larger subassembly components are costly to purchase,
the alternative, an in-house development effort, also requires a substantial
investment. It may be less expensive to purchase the component and integrate
it than to design, refine, validate, document, and produce a large subassembly
internally. The deciding factor will probably be the anticipated instrument
production volume.
System Control. The evaluation of controllers should be performed with care
to ensure that the processing power and interface options necessary to support
the project are all present. Without proper planning, such evaluation may not
be sufficient, which could lead to significant delays in the project schedule
if problems should arise. It is once again important to take a system view to
ensure that all disciplines are supported, including testing and manufacturing.
Today’s embedded micro markets provide full-featured processors with built
in digital I/O, A/D, D/A, and memory, all for $10 in large quantities. Suppliers
provide the development tools at no or low cost to accelerate the development.
When implementing the design, test points can very easily be included on a circuit
board if the need is considered in the design phase. The time involved in adding
them later and returning the board through the fabrication process during the
manufacturing stage can hold up the schedule by weeks. Such test points also
facilitate the software development effort when used in conjunction with the
rich feature set that the development tools provide.
A similar line of thinking should apply in dealing with component vendors. A
vendor is not going to rework its production line to accommodate a manufacturer
without foresight. Ensuring at an early project stage that the electrical components
have the necessary test points can save a lot of time, preserve good relations
with the vendor, and result in a successful product launch on schedule.
The Right Integration Mix. The cost of development, the cost of production,
and time to market are considerations always in tension and in balance. Similarly,
design and development is an iterative process, involving periodic reference
to time to market, system production cost, a possibly shifting sales forecast,
the dedication of effort to system integration, and the availability of mechanical,
electronic, and software components that suit the system architecture.
With every component selection the entire IVD system and its development project
should be taken into account. A component that may be the best choice from an
electromechanical standpoint could cost months in software development. It is
of little value to select a processor that has a six-month lead time. Every
instrument development effort is different, but each involves a mix of electronic,
mechanical, and software elements. The challenge is to find the mix best suited
to the system design that will demand the least complicated and laborious integration
effort during the design phase, prototype verification, and transfer of the
design to manufacturing.
While the architecture of the system may make writing custom software seem highly
desirable, customization may necessitate substantial validation and verification.
Off-the-shelf software also requires validation and verification when it is
integrated into a product, but the likelihood of bugs should be significantly
lower as long as the product is fully developed and reputable.
As the design effort iterates, the selection of components will change. A component
that had been eliminated as a candidate may come back into favor. Here is an
area in which regular communication with knowledgeable vendors can enhance the
development effort.
User Interface. From the operator’s perspective, the IVD system is often
a product developed on top of a Microsoft Windows–based platform. In addition
to the typical Microsoft tools for development such as Visual Studio, numerous
libraries exist to support all aspects of the development process. These include
cross-platform development tools, medical connectivity tools for HL7, libraries
for specific I/O boards, and dynamic link libraries for bar code readers.
Bar code readers are generally of the 1-D or 2-D type. The 2-D reader is gaining
in popularity because it can store significantly more information. Unfortunately,
information cannot be written back to the bar code. Alternatives involve radio-frequency
identification (RFID), which allows writing back to the disposable identifier.
The volume of information that can be stored ranges from bytes to kilobytes,
significantly more than with a bar code. Such a feature provides great flexibility,
but with a price penalty—one that may be worth paying depending on the
application.
Bar code readers and RFID readers have a similar look and feel. The read range
of RFID technology is typically a bit longer. RFID readers are often significantly
less expensive as well. Such technology is not the best solution for all applications,
but may be the answer for a specific need.
Working with Vendors
Time to market can be vital to a project’s success. Certainly, delays in
delivery of any component do not help. The component selection process calls
for attention to more than just the technical features and capabilities of a
subsystem; it is also important to ascertain vendors’ logistical capabilities
to supply prototype and production components when and in the quantities needed.
Paying visits to key vendors to get a first-hand view of their operations is
well worth the cost. Also important is knowing the vendor’s track record
of on-time delivery. That history may not need to be exemplary to ensure that
the vendor-customer relationship is successful. Knowledge and communication
and the best component are the cornerstones of a good relationship; a spotty
delivery schedule may be something for which the manufacturer can plan accordingly.
Asking the vendor for on-time delivery history, out-of-box failures, shipment
linearity, and other similar metrics is a key indicator of the commitment that
a supplier is making to ensure customer satisfaction. They demonstrate a vendor’s
desire to track key indicators of customer satisfaction and not simply revenue
generated.
Keeping current with technology is a major challenge for every company. Many
organizations focus on a core technology and outsource that which they see as
not central to their business. For example, a company whose core technology
is microarrays, FISH, or biochips may find instrument design to be beyond its
core competency. To fully comprehend the disciplines of instrument design, development,
QSR compliance, lean-flow manufacturing, microfluidics manufacturing, flex-circuit
assembly, ESD compliance, CE marking, and industrial design—let alone maintaining
a balance among their demands and requirements—can be too much for many
companies to expect of themselves. Since managing all these resources is at
times overwhelming, outsourcing enterprises that perform these services as their
primary business can be a great help to IVD manufacturers. (See the introduction
to Section 8, “Contract Manufacturing,” for a discussion of these
organizations.)
Both contract engineering firms and IVD manufacturers with their own engineering
resources should find this year’s IVD Technology buyers guide of great
assistance in selecting components for their systems. Both supply-chain managers
and engineers can identify and distinguish key vendors in relevant technology
areas. The guide is intended to save them time in their search for essential
IVD system components, and to make it easier for them to locate the right vendors.
Rick Muller,
HEI Inc., Advanced Medical Operations (Boulder, CO)
Copyright ©2004 IVD Technology
