The Department of Homeland Security is seeking to upgrade its biological sensor
network with more encompassing and less costly systems. The current program,
called Biowatch, has been established in more than 30 U.S. cities, but limits
on existing technology makes the process slow, dependent on human analysis and
costly to implement on a wide scale.
“Biowatch did what it was intended to do,” said Jane Alexander,
head of chemical and biological programs at DHS’ science and technology
directorate. “We went from a very slow way of knowing something happened
to knowing the very day it happens. It was a big leap forward in capabilities.
But I’d like to free up that money for the next important thing.”
In April, the Homeland Security Advanced Research Projects Agency (HSARPA)
selected 14 contractors for an 18-month $48 million research and development
project. They were asked to create a low-cost, reliable system that can identify
scores of pathogens in real time.
Half the contractors will work on “bio-agent autonomous networked detectors,”
expected to replace the Biowatch equipment for use in outdoor urban areas. The
other half is responsible to develop an indoor monitor called “rapid automated
biological identification system.”
Detector designers will try to lower costs by minimizing the size of the test
samples. Air samples currently tested at micro-liter sizes must be condensed
into nano-liters. Potential agents must be separated from background clutter.
DHS asked for a 2 cubic-foot detector that can operate without user intervention
for 30 days, testing the air eight times a day. The threshold for false alarms
is low, officials said.
“What is wanted is beyond the state-of-the-art,” Alexander said.
“The (contractor) community is very creative. If you don’t challenge
them, they won’t stretch … People surprised us on how far they could
go.”
Acquisition costs would run about $25,000 per unit, ordered in quantities of
1,000. Operation costs, including maintenance, spare parts and any consumables,
would be no more than $10,000 per system annually. The target price sought,
contractors told National Defense, would be an average of about 17 cents per
test, to check for as many pathogens as possible.
In comparison, current Biowatch systems cost $25,000, and roughly $2 to $3
per test, with each test only able to identify a single pathogen. If a detector
is to scan for a wide variety of threats every three hours, the cost quickly
soars. The Bush administration is seeking to increase funding for Biowatch from
$53 million in 2004 to $118 million for fiscal year 2005. The funds would be
used to set up additional sensors in high-profile areas.
The performance of the detectors, meanwhile, is limited by the available technology.
Samples from Biowatch sensors must be brought by hand to laboratories for testing,
leading to delays of more than a day.
“Something completely autonomous that delivers specificity simply doesn’t
exist,” said Duane Lindner, deputy director of Sandia National Laboratory’s
chemical and biological programs. “Collector-based systems (like those
used for Biowatch) are very specific and have low false alarm rates, but they
take a lot of time. When you start processing reagents in a 24/7 operation the
prices really add up to an astronomical level.”
One of the contractors working on the outdoor detection system said that DHS
requirements generally are more stringent than those in military bio-defense
programs.
The Army’s Biological Integrated Detection System, for example, can measure
the particles in an air sample to determine if an agent is present. BIDS also
is equipped with immunological assays that use reagent tests to mimic the human
body’s immune system, and can identify at least 10 agents in five minutes.
However, the system requires substantial power and hands-on laboratory preparatory
time.
David Hoey, vice president of business development for U.S. Genomics, said
that military customers usually train around the shortcomings of a system, such
as false positive results or complicated analysis, while civilian uses require
a more foolproof approach. Only a wide sensor net can protect civilian populations,
which in turn requires inexpensive and low-maintenance systems.
Biological agents come in many shapes and sizes, from delicate strands of filo
viruses to robust spores of anthrax. Furthermore, similar genetic profiles of
harmful and harmless microorganisms confuse many detectors, making them unreliable.
“It’s a big leap from prototype to an actual product that works
reliably, ” said Phil Belgrader, chief science officer for MicroFluidics
Systems Inc., another company that was awarded a DHS contract. Company researchers
are examining the best methods of condensing environmental samples into small
enough amounts to lower the cost of testing. “You want to keep reaction
volume low, because that’s where the costs are,” Belgrader said.
But increasing the numbers of agents the system can identify correctly is a
real challenge. DHS asked for the simultaneous analysis of a minimum of 20 pathogens
from the Centers for Disease Control’s Category A and B agent list, plus
scores of additional toxins and viruses. “We’re still looking at
the best approach to doing that,” Belgrader said.
Unlike chemical detection science, which has been steadily evolving since the
1920s, biological detection relied on growing cultures in Petrie dishes until
the emergence of DNA recognition technologies in the 1970s.
“One a scale of 1 to 100, chemical detectors are now at about 75,”
said Margaret Kosal, a staff scientist with the Center for Nonproliferation
Studies. “By comparison, because they’re about 50 years behind,
biological detectors are at about 25.”
Experts such as Kosel believe that layered systems are more effective. “There’s
not any one diagnostic tool that will do the job,” she said. “It
absolutely must be a combination of techniques. There’s no single silver
bullet.”
U.S. Genomics is proposing a system that does not rely on reagent reaction
but a pattern-recognition library of the DNA profiles of dangerous pathogens.
According to Hoey, the approach is akin to scanning a bar code at a supermarket
checkout. Instead of using tests customized to identify a single pathogen, as
is done with polymerase chain reaction testing, U.S. Genomics’ design
creates a library of linear DNA matches and uses a single test to identify many
agents.
The difference, put in supermarket checkout terms, is testing a candy bar with
specific equipment designed to determine if it’s a Snickers, and another
piece of equipment to determine if it’s a Milky Way, and so on, as opposed
to scanning the candy just once and finding it’s a Peppermint Patty. The
lack of narrowly targeted testing keeps the cost down, Hoey said.
Despite the difficulties in making the technology both reliable and affordable,
experts and researches alike say the creation of the ideal, round-the-clock
biological detector is feasible.
“It absolutely can be done,” said David Siegrist, a research fellow
studying biological detection for the Potomac Institute for Policy Studies.
“It can take a lot of money, more than we want, but if it doesn’t
violate the laws of physics, all we need are more engineers working on it. It’s
a grand scientific challenge.”