Copyright (c) 2005, Alvenh Channe
British Columbia Institute of Technology

In recent years, explosive sabotage by terrorists on transportation systems, buildings and crowds of people have claimed thousands of lives worldwide.  Because bombs and explosives that are capable of destroying aircrafts in flight can be very small and difficult to detect (Bryson, 1996), there is a heightened risk of terrorism.  Thus, many different technologies have been developed to assist security in detection of explosives.  Currently, a variety of screening techniques, from x-rays to trace detection systems capable of detecting minute explosive vapours are employed in airports as well as other settings such border crossings and court rooms (Connelly et al., 1998).

Despite advances in "instrumental sniffer" systems for bomb/explosive detection, canine/handler detection teams, one of the most primitive methods of detection, continue play an important role in detection.  Because the sensitivity of the canine (dog) olfactory system is "…equal or superior to machine technology in the areas of sensitivity and selectivity…" (Connelly et al., 1998; pp. 56), canines may be used alone in place of or in conjunction with "instrumental sniffers" (Connelly et al., 1998).  In addition, canines offer many other advantages over instrumental systems.

Unfortunately, there are many drawbacks of canine detection teams.  Because of size and cost, to mention only a few disadvantages, researchers (i.e. Otto et al., 2002; APOPO, 2003) have recently turned their attention to rodents (rats) as detectors of explosives and other contraband substances.  Although rodents offer many advantages over dogs (APOPO, 2003; Otto et al., 2002; Copley, 2002), should security managers deploy the use of rats in high security settings (i.e. court houses, borders, cargo, etc.)?  Can rats be used in lieu of canines and/or "instrumental sniffers"?

Since the use of rodents is a relatively new concept, security managers should be aware of any advantages/drawbacks if they are to considered for use in high-security settings, how they compare to dogs in terms of performance and reliability, and pertinent research finding.  Thus, the remainder of this study will compare dogs and rats in the detection of explosives and examine their advantages and disadvantages.

Introduction to Canines

For many years, the canine (dog) has been an invaluable aide in the search and detection of dangerous material.  While canines have been used for decades in landmine detection (Mott, 2004), the US Air Force first demonstrated the canine's ability to detect bombs, explosives, and illegal drugs in 1971 (Lemish, 1996, as cited in Connelly et al., 1998).  Since then, many military and law enforcement agencies worldwide have successfully employed canines to detect various dangerous substances (Johnston, 1999).

The dog's sharp sense of smell underlies the fact that canine/handler teams remains the instrument of choice for the military, government agencies and law enforcement agencies throughout the US (Connelly et al., 1998) in explosive detection, and even in the face of rapid technological advancements in the development of instrumental detection systems, there is a recent proliferation in the use of dogs (Seuter, 2003) in light of today's heightened risk of terrorism.  Brought about by many years of evolution, the dog's smell easily outperforms the most sophisticated instruments.  They are able to detect target odours instantaneously (as compared to machines), even in the presence of distracting smells (Johnston, 1999), and when canines pick-up target odours, they are able to follow its concentration gradient to the source (another task which machines are unable to perform) (Johnston, 1999; Connelly et al., 1998).

Using these abilities, explosive and bomb detection dogs are taught to recognize and respond to different vapours or molecules emanating from a variety of common explosives (Connelly et al., 1998) in addition to some rarer "European-type" explosives (Bryson, 1996).  Dogs are either trained on small quantities of actual explosives (Connelly et al., 1998) or with "dummy bombs" impregnated with harmless traces of parent explosives (Hunter, 1997).  Although, there are over 19,000 odours that are known to be associated with explosives, most of these odours are groups into categories (i.e. nitrates, acid salts, chlorates, etc.) from which samples representing these families are selected for canine training (Seuter, 2003).  It is assumed that canines trained on samples can generalize their ability to detect similar compounds within categories (Seuter, 2003; Connelly et al., 1998).  The Secret Service, for instance, trains canines on twelve different explosives (Bryson, 1998).  When explosives are detected, dogs are trained to give the same passive response by sitting and looking at the source while waiting for a reward (Connelly et al., 1998; Bryson, 1996).

Although bomb dogs are usually German shepherds, Belgium Malinois (Connelly et al., 1998; Bryson, 1996), or Labrador retrievers (Bryson, 1996) there are no "standardized breeds" specified for many K-9 explosives detection programs (Connelly et al., 1998; Bryson, 1996); therefore, other breeds may be used (i.e. hound dogs are also employed in security and police services) if they are motivated, can respond quickly, and are able to move around seats and containers in transportation search (Bryson, 1996).

Bomb sniffing dogs can be deployed in a variety of settings in which heightened security measures are required or when threats of explosives/bombs are received.  Such locations include, but are not limited to airports, cargo, mail, building, clandestine labs, and public areas (i.e. city parks, shopping centres, etc.) (Bryson, 1996), just to name a few.  Military dogs have also been deployed to sniff out landmines in war-torn countries (Mott, 2004).  As compared to machines, the higher quality "instrumental sniffers" tend to be the least portable (Connelly et al., 1998).

Canines also have the advantage of being cross-trainable for other tasks (Seuter, 2003; Connelly, 1998).  For instance dogs may be trained as patrol dogs or as detectors of drugs and alcohol in addition to their roles as explosive/bomb detectors.  Dual purpose canines can be a cost-saving measure (Seuter, 2003).  There are, however, conflicting views pertaining to cross-training of canines.  According to Seuter (2003), dogs that are cross trained for patrol are less likely to bite.  On the other hand, Seuter (2003) also suggests that cross training dogs to recognize different odours (i.e. narcotics, explosives, etc.) as a cost cutting measure "…is considered to be reckless by most professionals…" (Sueter, 2003) as it is unreliable, and can lead to confusion.  In addition, the FAA, suggests that for bomb detection dogs to be reliable, they must not be cross trained on narcotics and other substances (Bryson, 1998).

Limitations of Canines

Security managers should be aware of the many drawbacks pertaining to canine/handler teams.  Unlike machines and "instrumental sniffers", canines are live beings.  Thus, as with humans, they are unable to work continuously without rest, sleep, and eating.  The canine's ability to work is also affected by temperature and humidity (Seuter, 2003).  Canine/handler teams require a great deal of commitment, and dog must also be accompanied by specific human handlers (Otto et al., 2002) in order to work properly.  Machines, on the other hand, can work continuously and may be operated by any security personnel who are trained to do so, and thus, are more cost efficient (Connelly et al., 1998).  In addition, dogs are instructed to give the same alert for each explosive (Bryson, 1996), handlers are unaware of the exact type of explosives detected.  "Instrumental sniffers", on the other hand, are able to notify the user of the type of chemical detected.

The area to be searched must also be considered since, as suggested by Seuter (2003), height and size of the area can affect their abilities to search.  Settles and Kester (2001) further notes, dogs need to be in close proximity to sources in order to sample the scent, therefore it may be necessary to deploy multiple teams for larger areas as it may take awhile to sweep.  However, there is a possibility that teams may interfere with one another.

The issue of proximity is also of concern.  Dogs always sniff at distances of less than 10 cm and will move-in to essentially zero (Settles & Kester, 2001) as they follow a concentration gradient towards the source (Johnston, 1999) unless held back (Settles & Kester, 2001), and this presents some problems.  Getting very close to the source is dangerous if there are pressure sensitive explosives devices.

For smaller areas, accessibility is another issue of which security managers must be aware.  If dogs are unable reach into tight spaces, there is a danger of missing potential hiding spots.  This is, perhaps one of the most serious limitations of canine detectors; thus it is important for security managers to keep in mind that areas that have been swept by canines are not necessarily free of explosives or bombs (Seuter, 2003).

Introduction to Rodents

Given the limitations of canines, researchers are now studying the use of other animals in detection of explosives and other contraband material.  Animals such as bees and rodents (rats) (Mott, 2004) are currently being considered as replacements for canines.  Although, the use of rodents (and other animals) in detection is a relatively new concept, rats are already being deployed as explosive detectors in Mozambique, an African country scattered with landmines (Mott, 2004; APOPOS, 2003).

The basic underlying principle is similar to that of canines: rodents use their olfactory system to detect explosives and contraband substances, and alert for reward (APOPO, 2003; Otto et al., 2002; Copley, 2002).  In a study by Otto et al. (2002), rat alert by standing on their hind legs when target odours are detected, triggers computer sensors, which in turn notifies the handler and food reward is dispensed.

Rats are useful in detection because they offer many advantages over dogs.  One obvious advantage is that rats are cheap to obtain and easy to breed and maintain.  Female rats can produce litters (up to 10 pups in size) every few months (Otto et al., 2002), therefore it is possible to breed selectively for certain abilities.  In addition, the relatively low cost of training and maintenance (Otto et al., 2002) makes the rat an attractive alternative to its canine competitors.  Because rats can be trained using computer controlled instruments; "…it may be possible to train large numbers of rats in parallel and in an automated fashion" (Otto et al., 2002; pp. 218).  Unlike dogs, rats are not taught obedience, therefore, training period is relatively short (APOPO, 2003).

Because rats do not for social relationships with people (Copley, 2002), they do not rely on specific handlers.  Furthermore, rats can perform repetitive tasks for long periods (APOPO, 2003; Otto et al., 2002) and can be transferred to other handlers, allowing security personnel change shifts without interfering with search operation.

Rats are also easy to feed as they can live off food rewards (APOPO, 2003).  In the APOPO project, reward consist of banana and peanuts, and during the week-end, they are fed cereals, fruits and vegetables with some meat.

The small size of the rat offers another advantage over canines.  As previously mentioned, small spaces can be problematic for dog teams if dogs are unable to access potential hiding spots.  Rat, on the other hand, can squeeze into these areas for inspection.  While dogs need to be in close proximity to the source in order to detect explosives (Settles, 2003), rats, because their noses are close to ground in any position, are always at the point of highest vapour concentration in detection of landmines (APOPO, 2003).  In addition, their light weight makes it less likely to set off explosives (APOPO, 2003).

With respect to the rodent's sense of smell, can they match or outperform their canine competitors?  Rats have a highly developed sense of smell (Otto et al., 2002), and so far, research (Otto et al., 2002; APOPO, 2003) illustrating its efficacy in detection have been encouraging.  The rat's olfactory ability was demonstrated by Otto and colleagues (2002), in which animals (rats) were trained to sniff-out explosives and cocaine.  Rats were performed with a hit rate of greater than 90%.  In addition, the presence of distracter odours, such as engine oil and almond extract, did not seem to interfere with the rat's performance (Otto et al., 2002).

The rat's olfactory capabilities were further demonstrated in a study by APOPO (2003) that tested the detection threshold of rats for TNT; rodents showed reliable detection for concentrations as low as 0.001 ppt.  Currently, it is difficult to compare the limits of sensitivity to that of canines given that different studies with dogs (i.e. Williams et al., 1997a, 1997b; Johnston, 1999) seem to suggest different levels of sensitivity for canines, and explosives used were different from those used in studies involving rats.  According to Otto et al. (2002), rats have extremely poor sight and in the wild, they depend on their olfaction for survival, thus it can be assumed that the rat's ability to detect odours is superior to that of the canine.

Limitations of Rats

While properly trained rats can solve some of the problems associated with explosive detection dogs, security managers need to be aware of possible limitations.  Rats trained on a variety of chemicals can be useful in detecting different types of explosives; however, as with dogs, other than the fact the something dangerous is present, rats are not able to tell the handler whether smokeless powder or RDX was detected.

Unlike dogs, selection breed of rats should be considered carefully.  In the APOPO project (2003), an African breed of rat (Cricetomys gambianus) is selected for use in Africa, since they are less vulnerable to local diseases than imported breeds.
 
Suggestions for Further Research

Despite the current findings on the olfactory system of the rat, there is currently no conclusive data regarding its reliability as bomb/explosive detectors in airports, crowd searches and other settings outside the laboratory or mine field.  Further studies in these issues could yield more information that may be of help to security managers in considering the use of rodents.
 
Given that the rodent is one of the largest orders of mammals (APOPO, 2003) with many species of rats, more studies should be conducted before generalizing the abilities of the rats that have been studies to other species.

Conclusion

Despite advantages of rodents, one should be cautious in generalizing current findings to the context airport and other settings in which they are to be deployed.  However, given the limitations of canines, there is no reason why rodents, canines and instruments cannot be used to assist one another.  For instance, rats could sniff-out explosives in tight spaces that are otherwise inaccessible to canines, or are too dangerous for canine access.  In this day and age of heightened security measures in various settings, it is recommended that screeners use all the tools that are available to ensure public safety.

References

APOPO. (2003).  Frequently asked questions.  APOPO vapour detection technology.  Retrieved from http://www.apopo.org

Bryson, S. (1996).  Police dog tactics.  New York, NY: McGraw-Hill, Inc., College Custom Series.

Connelly, JM., Curby, WA., Fox, FT., & Hallowell, SF. (1998).  Detection of hidden explosives.  In A. Beveridge (Ed.), Forensic investigation of explosions. (pp. 45-71).  London, UK: Taylor & Francis.

Copley, J., (2002).  Rodents to rat on criminals.  New Scientist, Vol. 174(2347).

Hunter, S. (1997).  Taming explosives for training.  Science and Technology Review. UCRL-52000-97-9, 24-26.

Johnston, JM. (1999).  Canine detection capabilities: Operational implications of recent R & D findings.  Auburn, AL: Institute for Biological Detection Systems, Auburn University.

Lemesh, MG. (1996).  War dogs: Canines in combat.  Washington, DC: Brassey's Inc.

Mott, M. (2004).  Bees, Giant African Rats Used to Sniff Landmines.  National geographic.  Retrieved from http://news.nationalgeographic.com/news/2004/02/0210_040210_minerats.html

Otto, J., Brown, MF., & Long., W. (2002).  Training rats to search and alert on contraband odors.  Applied Animal Behaviour Science, Vol. 77, 217-232.

Settles, GS., & Kester, DA. (2001).  Aerodynamic sampling for landmine trace detection.  SPIE Aerosense, Vol. 4394(102).

Seuter, EJ. (2003).  The dog days of detection.  ECI: Published Articles.  Retrieved from http://www.nobombs.net/dog_days_of_detection.shtml

Williams, M., Johnston, JM., Waggoner, P., Jackson, J., Jones, M., Boussom, T., & Hallowell, SF. (1997a).  Determination of the canine detection odor signature for NG smokeless powder.  Proceedings of the Second Explosives Detection Technology Symposium and Aviation Security Technological Conference, Atlantic City, NJ: Federal Aviation Administration.

Williams, M., Johnston, JM., Waggoner, P., Jackson, J., Jones, M., Boussom, T., Hallowell, SF. & Petrousky, JA. (1997b).  Determination of the canine detection odor detection signature for a selected nitroglycerin-based smokeless powder.  Proceedings of the 13th Annual Security Technology Symposium and Exhibition Government. Industry Exchange, Virginia Beach, VA: American Defense Preparedness Association.

Other Research
Home