Some chemical agents, such as mustard gas, for example, have symptoms that appear much more slowly (NRC, 1999). Unconventional agents (e.g., aflatoxins, which can induce cancer in some exposed individuals) might not have observable effects for years. Effective responses to chemical attacks, and to biological attacks as well, need to be tailored to the specific agent involved; thus the choice of the right sensor(s) for the job at hand—whether with respect to the time scale or to other factors—is crucial.

In the United States, government-supported research on sensors is now mainly funded through DOD (DARPA), NSF, and DOE and has produced some significant advances in the sensitivity and other characteristics of the sensors themselves. In particular, sensors with medical applications have reached the market fairly rapidly, even though the DOD programs have focused mainly on military problems (e.g., standoff and point detection in field operations) and military customers. Development of sensors is heavily supported by industry as well, and industrial production facilities are routinely equipped with instrumentation that can detect and identify releases of toxic materials. However, none of these technologies has had any real impact on emergency preparedness, as the market for such applications is small and fragmented. For sensors to be effectively implemented for homeland security, they will need to be inexpensive, widely deployed, and networked.

Thus, although improved detection does not rely on sensors alone, research on sensors being conducted by many agencies, companies, and universities—including, but not limited to, work on sensors to detect explosives—should certainly continue. There are rich opportunities for discovering new technological principles on which sensors might be based.

Recommendation 4.1: A broad-based research program should continue to look for promising new principles on which better sensors might be based.

Presently, trained dogs represent the best broad-spectrum, high-sensitivity sensory systems. Dogs are capable of detecting many more items of interest, including people, explosives, drugs, fuels, and disease, and at lower concentrations than currently manufactured sensors. But the precise chemical signals that provoke responses in dogs remain uncertain; it is likely that the signals are not from a single compound but rather from multiple compounds. In the short term, the use of dogs could be expanded, and dogs could be trained to detect a wider array of targets. In the longer term, however, detailed studies to better understand the abilities of dogs could be useful in designing more broadly effective manufactured sensor systems.

Recommendation 4.2: Basic research to study how animals accomplish both detection and identification of trace chemicals should be pursued. These efforts could yield new concepts for better automated systems to reduce our dependence on the use of dogs for detection.



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