5

Final Thoughts

In the workshop’s final session, the session rapporteurs presented their reflections on the workshop. Michael Walter, Detection Branch chief and BioWatch program manager in the Office of Health Affairs at the Department of Homeland Security, also provided his concluding remarks. This chapter summarizes these comments and the ensuing general discussion among the workshop participants; Table 5-1 provides an overview of the four families of technology.1

THE PUBLIC HEALTH PERSPECTIVE

Beth Maldin Morgenthau, assistant commissioner for the Bureau of Policy, Community Resilience and Response within the Office of Emergency Preparedness and Response, New York City Department of Health and Mental Hygiene, said that the key message that came from the public health session was that when faced with a BioWatch Actionable Result (BAR), decision makers want more information as quickly as possible to help them put that BAR into context. Although each jurisdiction may have its own process for responding to a BAR and reaching a decision on what steps to take next, all of those processes benefit from having as much information as possible, whether it be from human and animal

____________

1The topics highlighted in this chapter are based on the summary of remarks discussed during each session. Additional comments by participants related to the closing remarks are also included. As noted in Chapter 1, comments included here should not be construed as reflecting any group consensus or endorsements by the Institute of Medicine or the National Research Council.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 81
5 Final Thoughts In the workshop’s final session, the session rapporteurs presented their reflections on the workshop. Michael Walter, Detection Branch chief and BioWatch program manager in the Office of Health Affairs at the Department of Homeland Security, also provided his concluding re- marks. This chapter summarizes these comments and the ensuing general discussion among the workshop participants; Table 5-1 provides an over- view of the four families of technology.1 THE PUBLIC HEALTH PERSPECTIVE Beth Maldin Morgenthau, assistant commissioner for the Bureau of Policy, Community Resilience and Response within the Office of Emer- gency Preparedness and Response, New York City Department of Health and Mental Hygiene, said that the key message that came from the public health session was that when faced with a BioWatch Actionable Result (BAR), decision makers want more information as quickly as possible to help them put that BAR into context. Although each jurisdiction may have its own process for responding to a BAR and reaching a decision on what steps to take next, all of those processes benefit from having as much information as possible, whether it be from human and animal 1 The topics highlighted in this chapter are based on the summary of remarks discussed during each session. Additional comments by participants related to the closing remarks are also included. As noted in Chapter 1, comments included here should not be con- strued as reflecting any group consensus or endorsements by the Institute of Medicine or the National Research Council. 81

OCR for page 81
82 TABLE 5-1 Potential Families of Technology for an Autonomous Detection System for BioWatch Nucleic Acid Immunoassays and Genomic Sequencing Mass Spectrometry Signatures Protein Signatures TRL of integrated systems using this TRL 9 TRL 6-plusa TRL 4b TRL 6 technology Noc Required Minimal Required • Less sample may Sample preparation • Cell lysis, nucleic acid • Cell lysis increases • Cell lysis, nucleic improve sensitivity extraction sensitivity acid extraction and specificity Nucleic acids, known Chemical Protein and structural Nucleic acids of and novel signatures epitopes, nucleic acids, targeted agents • Bacteria, viruses, • Bacteria, viruses, toxins of targeted • Bacteria, viruses, fungi, indirect fungi, toxins, agents Detectable agents fungi, indirect detection of toxins chemical agents • Bacteria, viruses, fungi, detection of toxins • Limited by • Limited by direct detection of toxins • Limited to included bioinformatic bioinformatic • Limited to included probes algorithms/ algorithms/ probes databases databases High High Moderate High Sensitivity • Enrichment • Single-copy detection • Some exceed current • May depend on (cfu/m3) strategies are needed for some organisms BioWatch requirements sample complexity for low-titer samples High Moderate Specificity High Moderate • Capable of • Not as high (false positive rate/ • False positive rate of • Near-neighbor single-nucleotide as nucleic acid speciation) 10–7 or better discrimination discrimination signatures

OCR for page 81
>1 hour <10 min (cycle times) Typically 3-10 minutes Processing time for • >10 hours Near real time • 4–6 hrs (collection, • Range: <5 minutes– detection (collection, prep, • Approx. 1 minute prep, and cycle times) 60 minutes cycle times, analysis) Inexpensive Moderate • <1 cent for each Expensive Relative operational Inexpensive • Consumables and analysis • Consumables and cost • Few consumables bioinformatic costs • <$100,000–$150,000 bioinformatic costs per unit Indoor/Outdoor Currently in clinical Operational • Less sensitive to dirty Indoor/Outdoor and research Indoor/Outdoor environment settings than nucleic laboratories acid–based technology Yes No No No Ability to test • Sensitive enough for • Used for ultimate • Used as • Used as trigger, with independently trigger, specific enough confirmation, trigger with PCR PCR confirmation for confirmation identification confirmation a These tests are commercially available and FDA-approved in some configurations, such as lateral flow assays. The “-plus” is intended to reflect this. b The lack of availability of an integrated system for sequencing is not simply due to technology limitations. No one has tried to engineer such a system due in large part to the newness of component subsystems. c Although smaller, less complex samples improve sensitivity and specificity, tiny samples reduce sensitivity, due to the effects of Poisson sampling errors for low-concentration agents. Specificity is reduced once the virus, bacteria, or fungus is fragmented by the ionization process. NOTE: cfu/m3 = colony-forming units per cubic meter of air; FDA= Food and Drug Administration; PCR = polymerase chain reaction; TRL = technology readiness level. 83

OCR for page 81
84 TECHNOLOGIES TO ENABLE AUTONOMOUS DETECTION FOR BIOWATCH surveillance, weather, past history, law enforcement information, or the results of other laboratory or field tests. The panelists from that session also noted that, in the end, their decision-making process is an art. “We will be making critical decisions without understanding the viability of the agent, the sensitivity to countermeasures, if this is an isolated inci- dent or part of a series of incidents, when and where the release took place, or how much has this been modified or weaponized,” Maldin said. “We are making these decisions without really understanding what the impact is to public health and just making the best guess we can.” This decision-making process becomes more complicated still when a BAR comes from an indoor BioWatch detector, which would trigger environmental sampling and require closing a facility. This is a potential- ly high-regret decision that instantly makes the BAR highly public at a time when an investigation is just beginning. If BioWatch plans to focus on indoor detection, it must do so in close cooperation not only with lo- cal public decision makers, but also with facility and transit officials, Maldin said. Along the same lines, Maldin noted that the panel stressed the im- portance of public health laboratorians and decision makers being com- fortable with whatever detection system BioWatch ultimately chooses. “We heard from the panel that there is no room for error. Public health’s currency is our credibility. We can’t make a mistake, and so we need to have complete confidence in the system.” Whatever system BioWatch chooses, she said, it must be the same or better than the current system in terms of sensitivity, specificity, and reproducibility. Desired Features of Future BioWatch Technology The main items on public health’s wish list for BioWatch were a shortened turnaround time to aid the decision making that must happen in the immediate aftermath of a BAR notification; higher spatial resolu- tion in terms of where an agent was released and at how many sites; characterization that would indicate whether an agent is pathogenic or not, viable or not, weaponized or not, and susceptible to medical coun- termeasures or not; and an estimate of how much agent was released. Also on the wish list were the ability to access an instrument remotely to obtain instrument-specific performance data that would help put a BAR in context, as well as the ability of the BioWatch detectors to store a sam- ple for additional testing and criminal investigation. Any future BioWatch

OCR for page 81
FINAL THOUGHTS 85 system should undergo extensive quality-assurance testing with the in- volvement of public health laboratorians, Maldin reiterated. Flexibility in being able to add or remove agents based on intelli- gence and scientific information is important, but so too is guidance on how public health can respond to any new agent, she said. “Don’t test unless you know what you are going to do with the results,” was an im- portant message noted by the panelists, Maldin said, as was public health’s need for the resources to respond to a threat. “If the alarm goes off and there is no one there to respond, then the system is worthless,” she said in commenting on the funding issues that most, if not all, public health departments have endured over the past number of years. For ex- ample, since 2005 Chicago has cut its public health staff from 1,600 to 600, while Texas has cut its staff from 700 to 500. Because the next- generation BioWatch system is going to be justifiably more complex and sophisticated, it will place an even greater burden on public health. “Not even the current resources will be able to support that,” Maldin said. “We need to think about how we are going to rebuild public health capacity.” In closing, Maldin said that the bottom line is that public health is in the business of saving lives and that it finds itself in the position of mak- ing critical decisions with little information. “If we get it wrong and overreact, the economic and political impacts of unnecessarily shutting down a facility or transportation are huge,” she said. “If we get it wrong and we underreact, the public health impact of doing nothing for those that have been exposed or delaying that are even bigger.” Public health is dedicated to getting the next-generation BioWatch system right and wants to be help design that system so that it provides the information that the end user—the public health decision maker—needs. NEXT-GENERATION DETECTION TECHNOLOGIES Nucleic-Acid Signatures William O’Neill, development program manager and project engi- neer, Biohazard Detection System, USPS, said that there are three systems—the microfluidic bioagent autonomous networked detector (M-BAND), the autonomous pathogen detection system (APDS), and the biohazard detection system (BDS)—that have provided effective aerosol collection, have demonstrated complex and automated sample prepara- tion, have performed multiplexed nucleic acid analysis, are TRL 8 or

OCR for page 81
86 TECHNOLOGIES TO ENABLE AUTONOMOUS DETECTION FOR BIOWATCH better, and have been in the field and running. All of these systems in- clude detailed reporting, automated chain-of-custody archiving, secure network communications, extensive diagnostic reporting, and detailed logistic support strategies. They also have various challenges in terms of incorporating new assays, given the expectations of zero false-positives from environmental samples, which will require extensive field testing prior to implementation, and there could be challenges regarding supply- chain issues if these systems are deployed extensively. The major challenge, according to the panel presentations and com- ments, will be the integration of multiple technologies from various ven- dors. In O’Neill’s view, it will be necessary to perform detailed systems engineering involving the technology developers to maximize system sensitivity in a cost-effective package. Two important technology chal- lenges, he said, are maximizing sample concentration and minimizing sample contamination in a system with high reliability. In order to meet this last requirement, he said, extensive and repeated field testing with detailed performance analysis will be a necessity. This can only occur as a sustained, multiyear effort, accompanied by a funding commitment from the BioWatch program, an idea that Eric Eisenstadt, independent consultant, noted as well. Although there are some issues with several of its assays, the real- time multiplexed polymerase chain reaction (PCR) has compiled an ex- tensive track record of accurate reporting. Thomas Slezak, assistant pro- gram leader, Informatics for the Global Security Program, Lawrence Livermore National Laboratory (LLNL), emphasized that even though BioWatch’s PCR-based system may have misidentified Francisella tularensis in field tests, that was a failure of the specific assay, not of the overall design of the BioWatch detection technology. The panelists agreed that PCR-based testing can serve as a primary detection assay and that false-positives can and must be minimized with orthogonal testing by any one of the many other analytical systems described by the panel. Besides their proven track record, PCR assays have well-established supply chains that can provide large quantities of the necessary reagents. One issue that must be addressed is sample preparation, particularly with regard to the ability to detect the species in differing environments, O’Neill said. Fluidics technologies may offer a solution to this problem. Another issue with PCR is the cost related to the licensing burden that comes with this technology. He noted, however, that the majority of the relevant patents will expire by 2016.

OCR for page 81
FINAL THOUGHTS 87 Immunoassays and Protein Signatures For immunoassay and protein technologies, the NG-ADS (next- generation autonomous detection system) array system is currently at TRL 6-plus and is capable of running immunoassays for rapid detection and nucleic acids for confirmation. What Thomas Slezak thought was important about NG-ADS is that this is largely an off-the-shelf system with an extensive product development history behind it. There was general agreement among the panelists that antibody-based detection is potentially fast but that it is not likely to achieve the sensitivity and specificity of nucleic acid–based technologies in a stand-alone next- generation BioWatch system, or at least not before 2020. However, be- cause of their cost, speed, ability to be automated and operate in a wide range of environments, and well-developed technology base, such sys- tems could find a role in combination with other technologies. Slezak added that given that these technologies all depend on recognizing sur- face epitopes, they will not be as good as genomic techniques at charac- terizing virulence and antibiotic resistance, which is another reason why antibody-based technologies are not likely to be used for primary detec- tion. However, protein-based assays should be better than nucleic acid technologies at detecting and characterizing toxins. Slezak stressed the importance of assay validation and cautioned that it should never be taken lightly. He added that the community should consider ideas for better validating highly multiplexed assays, and he said that protein-based technologies are undergoing field validation and are performing well, although much more extensive testing still lies ahead. He was particularly impressed by the ability of Raman-based sys- tems to detect both biological and chemical threats in the atmosphere, and he thought that the single-molecule array-based detection system had potential as well. One additional message that he heard from the panelists was the need for the Department of Homeland Security (DHS) to look beyond the familiar technologies and familiar development pathways. Genomic Sequencing One reason it is important to continue technology development even though PCR technology has proven to be a workhorse technology, O’Neill said, is the need to address the anticipated genetically engineered threats. O’Neill and Eisenstadt both said they were impressed with the new genomic technologies, particularly metagenomics and pan-genomics;

OCR for page 81
88 TECHNOLOGIES TO ENABLE AUTONOMOUS DETECTION FOR BIOWATCH these technologies can potentially address this issue with automated processing systems as previously discussed. Both O’Neill and Eisenstadt also said that they were encouraged by the significant effort being con- ducted to sequence large numbers of microorganisms and that they were intrigued by the idea of using nanopores to provide direct, reagent-free sequencing of a genome. O’Neill voiced the concern that nucleic acid technologies could produce so much information that it would add many levels of complexity that public health decision makers would then have to absorb. He expressed the opinion that the biodefense community needs to devote time and energy to develop guidelines that can help pub- lic health deal with this information. Eisenstadt agreed, reiterating how important it will be to involve laboratorians and public health decision makers early in the development process of these new nucleic-acid tech- nologies. He then noted that the real advantage of genomic sequencing technologies is that they provide the most complete and accurate view of the multiple organisms in the environment and the dynamic sequence space that they occupy. “The digital representation of the organism that you get via genome sequencing provides you with material that you can query to answer many questions, ranging from forensics to therapies to treatment to developing vaccines,” Eisenstadt said. Improvements in genomic sequencing, as is the case with the other characterization technologies that were discussed, will be driven by clin- ical applications, particularly for the point-of-care setting, Eisenstadt said. While the clinical opportunities are attracting funding that this bio- defense community can leverage, they are also raising technology awareness within the public health community. Genomic sequencing will not, however, be a panacea for biodetection, Eisenstadt said. “What is the proper role for genome sequencing in real-time or near real-time detec- tion when you are trying to monitor or identify when someone has delib- erately introduced a pathogen into your outdoor or indoor environment? We didn’t resolve that here.” It is still unclear, he added, whether genomic sequencing will be deployed in the field or will remain a laboratory-based system for confirmation and further elaboration of signals from primary detectors. Cost and training are potential challenges for genomic tech- nologies, and the field is evolving so quickly as to create a challenge for any acquisition programs such as BioWatch, particularly when it comes to validating the technology in the face of rapid change. Eisenstadt reiterated the need to develop automated and reliable sample preparation technologies. He also highlighted the need to con- struct and maintain high-quality, curated databases and to develop the

OCR for page 81
FINAL THOUGHTS 89 bioinformatics tools that will be needed to make sense of the large da- tasets that most of these new technologies, genomic and others, will gen- erate. As O’Neill discussed, integration is a challenge that needs to be addressed. Eisenstadt highlighted the development of aerosol-to-sample analysis technologies and the aerosol characterization processes as hav- ing great potential. He expressed concern that the BioWatch program needs to ensure that there is enough flexibility in the acquisition plan that it will not miss out on capitalizing on the genomics wave when it peaks. Mass Spectrometry According to Charles Kolb, president and chief executive officer, Aerodyne Research, Inc., a strength of mass spectrometry is its ability to characterize both targets and the background—an important capability because the issue BioWatch faces is not signal-to-noise but signal-to- background. One way to address that challenge is to concentrate the tar- get as much as possible and to reduce background content, while another is to understand the background as much as possible, and it is at this se- cond approach that mass spectrometry excels. Another strength of mass spectrometry is that it is supported by a huge technical base of instrument development. After reviewing the basic mass spectrometer components, Kolb said that the available technology affords a wide range of choices in terms of cost, size, maximum resolution, and other characteristics. He offered the opinion that while real-world mass spectrometry applications are evolving rapidly, it will take a number of years to settle on the best combinations of technology and data analytics for BioWatch and then field test the resulting instruments. Nonetheless, Kolb said, the panelists pointed out several areas in which mass spectrometry could play a role in BioWatch. As a real-time sentinel, mass spectrometry could provide a signal that would trigger more specific technologies as it detects suspicious particles. It could also provide background characteristics that could be used to correlate prob- lem particle types with wind direction, temperature, and diurnal and sea- sonal variations and to provide real-time data to allow back-dispersion modeling to characterize release and forward-dispersion modeling to predict impact areas. With further technology and database development, mass spectrometry may be able to improve detection specificity.

OCR for page 81
90 TECHNOLOGIES TO ENABLE AUTONOMOUS DETECTION FOR BIOWATCH THE LABORATORY PERSPECTIVE As a potential end user of the technology that had been discussed during the course of the workshop, Sandra Smole, director of the Divi- sion of Molecular Diagnostics and Virology in the Massachusetts De- partment of Public Health’s Bureau of Laboratory Science, said that she was excited about these technologies but she was “very concerned that many of them will not or do not meet the state of readiness that is re- quired for the BioWatch autonomous detector and, certainly, not the state of readiness required for deployment in the BioWatch system.” From an operational perspective, she said, the public health response ac- tions that are initiated as a result of these technologies must be supported by a robust system meeting all of the technical specifications that are listed in the DHS request for proposals as well as the needs of the public health laboratorians and decision makers. As a laboratorian she is most interested in understanding the performance characteristics of these sys- tems and then being able to network into and access these systems to monitor their performance. The most critical information needed from an autonomous system is the confident identification of the biological threat to inform follow-on actions and laboratory analyses. One of the most important roles of the laboratorian is to explain the results that are delivered to the public health decision makers. From that perspective, having multiple detection capabilities that can provide in- formation on drug resistance markers, viability, and virulence can be very useful. “As a laboratorian, I am very interested in strain typing and fingerprinting,” she said. “Public health laboratorians spend a considera- ble amount of effort in tracking down the source of infection and out- breaks, and I am very interested in applying technologies that allow for that capacity.” In particular, she noted the potential of mass spectrometry and genome sequencing to provide that type of information, but she also noted that developing and networking well-curated databases will be crit- ical for realizing that potential. In addition, she commented that this de- tailed characterization would not be needed immediately as part of an autonomous system, but could be provided by the lab-based investiga- tion. She also reiterated the comment made by several speakers that BioWatch can benefit from the efforts of the clinical diagnostics com- munity and the efforts that are going into developing products for those end users. As a final comment, Smole agreed with the general discussions that nucleic acid–based technologies are currently at the high technology

OCR for page 81
FINAL THOUGHTS 91 readiness levels, but she made a plea for the developers of other technol- ogies to continue their work in this field. “See if you can try to develop these technologies and bring them up to speed and compete,” she said, while acknowledging that such an effort will require funding that may be hard to procure. DISCUSSION Toby Merlin from the Centers for Disease Control and Prevention (CDC) started the discussion by asking the workshop attendees to think about the role that technology can play in managing risk, that is, by providing information that enables the decision maker to make the best possible decision in the face of uncertainty. In that regard, he noted the importance of high-quality, highly trusted data for the decision maker. Increasing the confidence that the decision maker has in the quality of the data that the technology provides lowers the decision maker’s risk. He also pointed out to the audience that while this workshop focused on aerosol release, there are many other avenues of release that BioWatch does not cover. Suzet McKinney, deputy commissioner for the Chicago Department of Public Health, said that she has been involved with BioWatch since 2005 and that this was the first time that she could remember public health having the opportunity to have a discussion with both DHS and industry at the same time. “Public health would like to be involved and would like to continue to be involved in this discussion,” she said. Speaking to the technology developers, she said that while the technolo- gy development community wants to test these technologies in an opera- tional environment, she and her public health colleagues do not want to see the results of those tests in real time. “If we see them during the test, we have to act on them,” she said. “Until we have the confidence in these technologies that the actions are sound based on the fact that the technol- ogy works, we are not going to want to go there.” Scott Hughes from the New York City Public Health Laboratory added that it is important to share the data from these tests eventually so that he and his colleagues can become familiar with the data and gain confidence in them. David Persse, emergency medical system director for the City of Houston, said that one piece of information that public health needs to makes its decision is how different a result is from the local environmen- tal background. He reminded the audience that when BioWatch was

OCR for page 81
92 TECHNOLOGIES TO ENABLE AUTONOMOUS DETECTION FOR BIOWATCH launched and had its first BAR, there was not a clear understanding of whether that result was due to the presence of microorganisms that natu- rally existed in the environment or pathogen purposely introduced into the environment. Persse said that context is an important factor for public health decision makers to consider; for him, this information is important when explaining decisions to his mayor. Erica Pan, director of the Divi- sion of Communicable Disease Control and Prevention in the Alameda County (California) Public Health Department, added that it is not just the mayor who needs to understand the reasons behind public health’s decision; the public does, too. Jeffrey Schloss from the National Human Genome Research Institute asked how feasible it would be to swap out modules in a deployed in- strument because it is likely that inserting a new detection technology would also require inserting a new sample-processing module. Another participant asked if there was some metric for how big an improvement would need to be to result in swapping out technology, given the time and cost involved in making such a transition. O’Neill answered the first question by saying that such flexibility is possible but that any transition would take 12 to 18 months, given the thousands of units that would have to be tested and validated and the training that would have to take place. Slezak agreed with that assessment. “Switching out a nucleic acid PCR-based system for any of the other wonderful technologies we heard here is ground zero,” he said. “You would have to build a whole new system from scratch and test the whole system in parallel with the exist- ing one.” Allen Northrup, chief executive officer of the Northrup Con- sulting Group, disagreed with this assessment, commenting that one way to improve modularity is to understand the basic operating parameters of the existing system in terms of inputs and outputs and to design replace- ment modules to fit those parameters from the beginning. William Raub, chair of the workshop planning committee, asked if the idea of combining two orthogonal testing methods in the same in- strument was too big a challenge in terms of integration and operating reliability. He wondered if a better approach might be to design a semi- autonomous system that would require a laboratorian to make the deci- sion whether to trigger a second detection method based on the results of the primary method. John Vitko, who moderated the first technology ses- sion on nucleic acid signatures, said he was skeptical about the practicali- ty of integrating two detection systems on the same deployable platform in part because of cost, but largely because of the difficulty of optimizing two completely different technologies to run in the same instrument.

OCR for page 81
FINAL THOUGHTS 93 THE BIOWATCH PERSPECTIVE This workshop, Michael Walter said, was incredibly important for the BioWatch program because it provided the opportunity for the bio- detection field to examine the whole range of technologies, from mature to developing, and to get a sense of the technology pipeline so that BioWatch program officials can make an informed decision on those areas that it should focus on. He also reminded the workshop participants of the complexity of the BioWatch program. “It has an enormous number of moving parts, and the ramifications of the decisions it makes are enormously important,” he said. “When we look at how we are going to move forward with the BioWatch program, it is important to remember that it has got to fit into the public health realm, and to do that it has to be better, faster, cheaper. Those are our three mantras that Congress as- signed to the program. That is what we need to look at.” At the same time, the BioWatch program cannot fund research and development—it is concerned solely with deploying an operational sys- tem that generates accurate, actionable results for use by the local public health jurisdictions. He noted that there was some concern voiced at the workshop about where information from the BioWatch detectors would go, and he stressed that this is a nonissue, as all information goes to the public health departments in the local jurisdictions. It may happen, he said, that CDC would get the information simultaneously, but he stated clearly that “the information will remain under the control of the state and local public health departments.” In commenting on the prime requirements for technology for BioWatch, Walter said that “the last thing we can do is test, get a result, test, get a result, and test again. We don’t have the time for it and we don’t have the money for it. Whatever we put out there has got to be fast, accurate, and has got to be right.” He went on to say that the program is not allowed a false alarm, for that could jeopardize the program, and he acknowledged the extreme scrutiny that the program is coming under from Congress, the White House, and DHS. He also remarked that the procurement phi- losophy for BioWatch holds that while it may deploy a particular tech- nology in an autonomous system, it will replace modules within that system as the technology improves and is validated in the field. As a re- sult, BioWatch will continue to be interested in technology innovations in the years ahead.

OCR for page 81