Inside the forensics world

[Professor Shari Forbes from the School of Mathematical and Physical Sciences speaks on stage]

Shari: Given what I‘m talking about, which is death and decomposition, I have to give the following caution to you all. The presentation does contain photos of decomposed animal and human remains and I simply ask that you use your own discretion as to whether or not you wish to view these images. Okay. With that out of the way, when I tell people what I do for research, the first question I’m always asked is ‘How do you actually study decomposition?’ And if we were in the medical sciences studying life and death, we would use the laboratory mouse, or the laboratory rat, or, for the science students in the audience, maybe even the frog. But in forensic science, we typically use the domestic pig. And we use that as an analogue for human decomposition. There’s several reasons why we do that. The main reason is that it we believe that human remains undergo the same process of decomposition as pig remains – namely, we believe they go from fresh through to bloat, active decay, advanced decay, and the final stage, which is skelotonisation. And although we have used the domestic pig for years, if not centuries, as an analogue for studying decomposition, we have never actually proven that they’re adequate or acceptable analogues for human remains. So many researchers say, or believe, that the only way to really study human decomposition is to use human cadavers, but that begs the question, can we ethically, morally and legally do that? And the answer is we can, but only in very select facilities. Those facilities are often referred to as body farms. Now, the term body farm was not coined by forensic scientists; it was coined by Patricia Cornwell when she wrote a novel of the same title in 1994. She wrote that novel after she visited a facility, a facility that at the time was the only facility in the world where people could donate their body to science for the purposes of studying decomposition. That facility was based at the University of Tennessee Knoxville, and was founded by this gentleman, Dr Bill Bass. And many of you may have heard of Dr Bass, because he is a world-renowned forensic anthropologist, which means he studies human remains, and particularly skeletal remains. Dr Bass is often consulted by police, and he’s asked to do usually two things: he’s asked first to provide a profile of skeletal remains, and that means age, gender, weight, height – different variables about the body. The second thing he’s often asked to do is to estimate time since death; that is, how long a person has been deceased. And it was during one particular case – in fact, it is the strange case of Colonel Shy – that Dr Bass realised the need for this facility. Dr Bass, this was in 1977, was called to an investigation in the early hours of the morning when the police discovered, and I’ll quote here, ‘the grisly remains of a young man, headless and clad in a black tuxedo, found on top of the grave of Civil War hero Colonel William Shy.’ And Colonel Shy had died during the Battle of Nashville. The remains were collected, they were sent to the lab of Dr Bass, and he was asked those two things: firstly to provide a profile of this individual. He told the police that this individual was approximately 26-29 years of age at the time of death, that he was a male Caucasian, that his height was about 5’11” and his weight was around 175 pounds. The police also asked ‘Is this a recent victim?’ And he said ‘I believe this victim has been deceased for about 2-6 months.’ Time since death, 2-6 months. So clearly this was a murder. And obviously the police started a homicide investigation. Unfortunately the police wasted much time and resources on this homicide investigation until the truth finally came out, and later Dr Bass was rather famously quoted as saying ‘Well, I got the age, sex, race, height and weight right. But I was off on the time since death by 113 years.’ Because in fact, these were the remains of Colonel William Shy – his body had been pulled from a very small hole in the coffin that the police had not discovered, and his grave had been robbed the night before. So why then were his remains so well preserved? Because in fact, he had been embalmed, which was traditional for that time when he died. And that was the reason that Dr Bass’s estimation was so inaccurate, but it was also the reason that Dr Bass said ‘We really need a facility to study decomposition and particularly to study preservation.’ And so that facility was opened in Tennessee in 1980; some of you may have heard of it. And for 26 years it was the only facility of its kind in the world. Since 2000, many more of these facilities have opened across the USA in very distinct different climactic regions. And that’s really important, because we know that the rate of decomposition varies based on the environment in which it occurs. So we know for example that a body decomposing in the snowy winter months of New York will be very different to a body decomposing in the hot, dry summers of Texas or Arizona, for example. And like Dr Bass, we know that we have to study the rate of decomposition in our own environment if we’re ever going to have a hope of estimating time since death accurately. We also know that here in Australia, our climate is very distinct from the USA. And if we look at global maps of the ecosystem – soil, vegetation, insects, other invertebrates and vertebrates, again we can see it’s very distinct to the US. And so that again posed the question to us, can we really extrapolate data from the US and apply it to forensic investigations here in Sydney, Australia, and can we do it accurately? And the answer is no, of course we can’t. So that’s the reason why 12 organisations, including the University of Technology Sydney recently opened what’s now known as Australia’s first body farm. It’s the only one in Australia, the only one in the Southern Hemisphere, the only one located outside of America in the world. We don’t like, however, to call it a body farm. So being academics, we gave it an extremely long and tedious name: the Australian Facility for Taphonomic Experimental Research. We just refer to it as AFTER. That was intentional.

[Laughter]

Like those facilities in the US, AFTER is a high security facility, as you can see by the prison-like fence, and it is only available to authorised users, because maintaining the privacy and dignity of our donors is our number one priority. Also like those facilities in America, once you get past the fence, and usually a very small building at the front, all that’s inside is natural bushland, because we’re trying to study the natural process of decomposition in an outdoor environment, and we’re trying to mimic the types of environments that the police may be searching for victim remains. So who are these victims that they’re helping? It could be missing persons, and not every person goes missing as a result of a crime, so it could be a bushwalker, somebody who’s gone hiking in the Blue Mountains, which is actually not far from our facility – they’ve unfortunately become lost, they’ve perished, and their remains now need to be recovered. But in many instances, unfortunately, victims can be victims as crime, such as the backpacker murders. Infamous and notorious serial killer Ivan Milat. More and more, our police are now being asked to assist with victim recovery following mass disasters, whether they’re natural, such as earthquakes and tsunamis, or whether they’re intentional, such as the German Wings air crash shown here, or terrorist attacks such as 9/11. Worse still, our forensic scientists are now involved in mass disasters following war crime – acts of genocide, where hundreds and hundreds of individuals are buried in mass graves. So these are the kinds of investigations that the police and forensic scientists are working in, but then why do we need to study death and decomposition? How can our research at AFTER possibly help these investigations? There’s a number of things we can do, but it’s mostly about enhancing the methods we have available, or that are provided to the police, methods that we can use to search and locate victim remains, one of the most challenging things to do. Methods such as cadaver detection dogs, and I’m going to come back to this later, because this is my area of research, and I’ll talk a little bit about what we do in my group. But there’s a lot of other search tools – geophysical search tools, such as ground penetrating radar, and if we’re searching very large areas, aerial search tools, things such as hyper spectral imaging, thermal imaging, and radar – all of these can now be used in forensic investigations. And once we locate a victim, then we need to recover those remains. And we apply archaeological recovery techniques to ensure that we’re gaining the most evidence from that scene. Not just the body, which of course is crucial, but evidence that can link the victim to the offender to the crime scene. Evidence such as tool marks in the grave walls that may have been left there by a digging implement, or the footwear impression of an offender, placed there when they were digging the grave, concealed by the body until we can remove them and discover that evidence. So it is extremely important that we use these tools. And once we recover victims, then we need to think about how long have they been there? Like Dr Bass, we strive to enhance our methods for determining time since death. And at the moment, for decomposed remains, the most accurate method we have is forensic entomology – studying the life cycle of insects, usually beetles, flies and their larvae, as shown in here. And it’s only as accurate as the methods that we can use and the environments that we study them in, so we have to be able to study these life cycles in our ecosystems here in Sydney, Australia, if we’re going to use it accurately in estimating time since death. Once we’ve recovered the remains, determined time since death, then the next thing to do is to identify them. At the moment, there’s three primary identifiers we typically use to confirm an individual’s identification. The first is fingerprints, and you’re going to hear about that next tonight. The second is DNA, and the third is teeth, compared to dental records. All of that assumes, though, that one or more of these identifiers are available at a crime scene, and when we’re dealing with crime scenes like this, the recovery area of the MH17 crash, it’s highly likely that those identifiers won’t be available. So our research then looks to other methods – what else can we develop to assist the police in identification? Things such as isotopic analysis – it won’t help to identify an individual, but it can tell us their geographic location during life; where they lived. And when we’re talking about an air crash with people from all over the world, that information can be very, very key. We also need to think about different contexts that the police would typically search in. What happens to the rate of decomposition when a body’s buried in a shallow grave, and how does that change if the grave is much deeper? We know that it actually slows the decomposition considerably. What happens if they’re concealed in some other manner? Bodies are often recovered from the trunks of vehicles. They can be recovered from mobile homes, and they’re regularly recovered from just normal homes. In fact, 80-90 per cent of victim recovery is in the average household in Australia, from indoor structures. So these are all environments that we need to study, because it changes the rate of decomposition and it changes our estimation of time since death. We also know from the police, and we work closely with them, that they are required to recover remains from water. And we know that when a body comes out of the water, it often contains this white substance known as adiposea, and that’s key, because adiposea can assist in preserving soft tissue, and perhaps even evidence of the crime. We also have to work in other environments such as fire. Whether it’s an intentionally set fire, which is arson, or whether more commonly it’s our natural bushfires, which are common every summer of every year here in Australia. It’s unique, really, to our environment, that we continue having to study what happens to bodies that are really involved in a fire and then subsequently decompose. So all of our research is about how can we search for, locate, recover and identify victim remains? Particularly in these kinds of scenarios when so little is still available. And so that’s what we do at AFTER – we constantly strive to improve those methods, and I would like to tell you a little bit of our results that we’ve got so far out there, but I have to mention AFTER only opened earlier this year. So we’ve only been open about seven months, and we don’t have a lot to share in terms of our results. So before I talk a little bit about our AFTER results, as I mentioned, I’d like to highlight our other area of research, on that we’ve been working on for years, and that still fits within that AFTER facility, and that’s the dogs. So we work with scent detection canines, specifically, or particularly, we work with cadaver detection dogs. And as beautiful as their pups are, and these are there pups, it’s these equally, in my mind, beautiful individuals that that we get to work with through a collaboration with the NSW Police Cadaver Dog Unit. In NSW, there’s only four cadaver dogs, and in all of Australia, there’s about the same number again. So there’s not many cadaver dogs because thankfully, there’s not a lot of need for them. The four that we get the opportunity to work with are from left to right here: Bronte, Chloe, Digger and recently graduated Bryson, who’s on the left. And you may have noticed that they’re predominantly English springer spaniels, and I often get asked, as do the police, why English springer spaniels? Why are they used so regularly for cadaver detection work? There’s a lot of different reasons why they use them, but there’s one main reason, and that’s because they have loads of energy – just loads of energy. And so that means they have really high drive and they will have an outstanding work ethic. They will work and work and work until exhaustion, if you let them. So they’re great for searching really large areas, and these are the kinds of areas that police may be searching for victim remains. But because of their agility, they’re equally as good at searching confined areas, and as I mentioned, these are also environments where police may be searching. And importantly, as I’ve mentioned, their work ethic is outstanding, and they will work until they get told to stop. And they do it for one thing: they do it for the opportunity to play ball with their handler. That’s what they do it for. They don’t understand what they’re doing; they just know what gets them their reward, and that is to recognise their target odour, and for cadaver detection dogs, that’s decomposition odour. We know, though, how they do it, so we already know that they use what’s known as a scent pool and a scent cone to locate that target odour. During life, we all have a unique odour profile, unique to ourselves and different to your neighbour or anyone else. But following death, our research has shown that that unique living scent becomes a more generic decomposition scent, and that’s the scent that stays around your body. With just the slightest bit of wind, that scent pool can instantly become a scent cone, and as long as the dogs can pick up even a few molecules on the edge of that cone, then you’ll see them track it back and forth, picking up the edge of the cone until they’ve reached the point. And that point is their target odour – it gets them their reward, and hopefully it shows us where the victim is. And so that’s what you can see here, in terms of the scent cone. So we know that that’s how they do it. What we don’t necessarily know is what is actually in that decomposition odour? And what are they actually recognising to get them that reward? And that’s where our research comes in. So we’re able to chemically profile the odour of decomposition and we’ve found that there are hundreds and hundreds of compounds. We’re able to show that it’s very dynamic – not only does it change throughout the decomposition stages, but it changes by the minutes, hours, days, weeks and months. We know that the dogs can detect decomposition odour across all of these stages, and with post-mortem intervals of minutes, weeks, days, months, whatever. So we suspect that there must be a few compounds in that odour profile that the dogs recognise, they remember and they know this is what gets me my tennis ball, and that’s how they do their work. We know this, because we’ve been able to show that chemically there is an odour from even very old remains. Skeletal remains that are more than 20 years post-mortem, and the dogs have successfully alerted on that. We’ve also been able to show that chemically, even when a body is no longer in an environment, there is residual scent left in that environment, so we can detect it and the dogs can detect it also, so that body may have been moved by an offender revisiting the crime scene, or more likely, by scavengers who have moved the remains elsewhere. But the dog can still identify that original deposition site, and then hopefully track to the other deposition site as well. We’ve also shown with our research that even when a body’s buried, the odour can permeate through the soil, and it can reach the grave’s surface. Now, it’s very dependent on burial depth, on Ph, soil type, moisture, a hundred different variables, but it can be done. And if it can reach the grave’s surface, then we know and we have shown that the dogs can detect it. So it’s often a query, can they really detect buried bodies? And the answer is they can, but it does require certain conditions. We also know that they have to work in these really challenging environments now that maybe they didn’t have to a while back. Mass disasters, such as this, where it’s really hard to know which dogs to send in. As I said, we have a unique living scent, but that transitions into a decomposition odour, but at what point does it change? Immediately after disaster, do we send in a live scent tracking dog, even if the individual’s deceased, or do we assume they smell like decomposition scent and send in the cadaver dogs. These are the kinds of questions that urban search and rescue teams ask us all the time, and ones that we’re trying o address, to just know the absolute best opportunity in terms of which dog to send, the likelihood of recovering living victims first, but then recovering deceased victims equally as fast. And one of the areas that we don’t work in in Australia that I’d love to start working in is water detection dogs. It’s true cadaver detection dogs can locate the odour of decomposition in water, just in the same way they can with burials. And you can imagine with the water environments surrounding Australia and within Australia, this kind of detection is really key. It’s one we don’t have, but it’s one that we’re looking to investigate in the future. One of the new ones that we do get to work with, though, this is a very new kind of detector dog in Australia – this is Sylvia. Sylvia is Australia’s first blood detection dog, and there’s only a few of them now in Australia. Sylvia is trained just to detect blood – she’s only trained on blood, she’s only looking for blood. That’s what gets her her reward. But when we’re talking about blood, we’re talking about really trace concentrations and we’re talking about latent blood – blood that’s not visible to us, but that she can still detect. And that’s really important at crime scenes, because we’re always looking for latent blood. We also look at what happens when you weather these things, so Sylvia’s so new there’s lots of questions we have – how long can she detect blood at a crime scene for? This is something we’re learning. What happens if the blood’s been exposed on concrete to rain, wind, sun, these kinds of things? Can we detect an odour, and can she? What if it’s a different surface? What if it’s wood? Can she detect an odour, and can we? And what happens when an offender tries to clean up, because often this happens with blood – they’ll try to clean the crime scene, or if its on their clothing then they’ll wash their clothing. So we can see blood in the top left image, and after washing, we can no longer see it. This is latent blood. Currently, we use luminol – this is our chemical detector that tells us it’s still there. The question is, can our biological detector – that is, Sylvia – detect it, and do so non-invasively, without the need for chemicals? So these are the kinds of questions that we’re still asking. We suspect Sylvia can detect it, because dogs are reported to have a sensitivity of 1 part per trillion – that is, one teaspoon of sugar in two Olympic sized pools. Now, as amazing as that sounds, our research has shown that the dogs actually have a far lower limit of detection than 1 part per trillion, so that’s why we believe Sylvia probably can detect latent blood. So we know that their olfactory system, these biological noses, will never really compare to the chemical or the electronic noses that we design. But it doesn’t mean we stop designing them – some of our research is about looking at chemical detectors, e-noses, as you can see here, one of the ones we’re designing at the moment. Because there are instances where we simply don’t want to send dogs – where it’s far too hazardous for them and their handler. And in those instances, having an electronic detector, even if it doesn’t have the same sensitivity level, is still important. It’s still valuable. So we don’t discount them altogether; we just recognise that probably in my lifetime, we’re never going to come up with chemical detector that’s as sensitive as the dogs’ nose. But the thing that all of our research has shown us, and the thing we constantly learn from working with these dogs, no matter what kind of detector dog they are, they’re only as good as their training. So we can’t ask them to do something that they haven’t been trained to do, and we can’t expect them to work in all environments if we don’t know the limitations. So a lot of our research is about understanding those limitations, knowing when they can work, and making sure they have the best training and the best training aids to ensure their success when they’re deployed to crime scenes or to disaster scenarios. So as I said, I do have a few slides that I’d like to share with you about AFTER. We do a lot of our cadaver dog research at AFTER, but there’s a few other things that we’ve been doing in the six months that we’ve been open, and the first and most obvious thing we did was to compare pigs and human remains. And perhaps surprisingly what we found in our distinct environment here in Sydney, Australia, is that pigs do not equal human remains. And perhaps it wasn’t that surprising, but we found that the rate of decomposition and the stages of decomposition are actually very different. Now, we’re not discounting the use of pig remains – we recognise their importance, but we understand now if we’re going to use them to estimate time since death, we have to factor in an error into that estimation, because with human remains, they don’t go through all those stages. They go through some of the stages, but ultimately they end in mummification. And they look not dissimilar to Utsi the iceman shown here. Now, Utsi the iceman, of course, was discovered 5300 years post-mortem, and we won’t be studying our research for that long, but we would like to know how long is a body preserved in our environment? And it’s really important information, because preserving a body in a natural environment can help to preserve some of those identifiers that I said are often not available to criminal investigators, so it is key that we know how long that information  may be available in terms of police investigations. And finally, the most interesting thing we’ve learnt, or at least personally I’ve learnt, is nothing to do with research, but it is related to AFTER. And it’s been about the reason that our donors have decided to sign up to AFTER. We have been inundated with people signing up to donate their bodies to AFTER, and I always ask people why they’re choosing to do so. And we hear lots of fabulous reasons – probably the most common is definitely altruism; we hear people say ‘My body is just a shell – once I die, I’m not using it, so somebody else might as well.’ And that’s one of the reasons that they sign up to science. Another reason we hear is that they are real supportive, very supportive, of environmental sustainability, and they like the idea of simply returning naturally to the earth. And so I think that’s a great example of why somebody might want to sign up to AFTER. Still others of our donors just have a really great awareness of what we do, whether they’re fans of CSI or if they’re fans of Patricia Cornwell, we don’t care, as long as they now what we do, they understand why we do it. They understand the need for these facilities and the impact that it can have on society. And ultimately the thing I continue to learn on a daily basis is the exact thing that Dr Bass knew all those years ago: that it doesn’t matter how much we study, and we can do this for decades more, there’s still so much we need to learn about the processes of decomposition and also preservation. Thank you very much.

[Applause]

[Dr Xanthe Spindler, Postdoctoral Research Fellow in the School of Mathematical and Physical Sciences speaks on stage]

Xanthe: So, to give you a little bit of context as to why fingerprints have been around for so long in forensic science and why they’re so fascinating to me and to my students and to hopefully a lot of you in the audience, we’re going back to the age of enlightenment. So when we talk about fingerprint detection from a historical and from a modern forensic science perspective, we’re really starting in the 1890s. Around 1891, we start to see the first research into fingerprint identification as a method of identifying criminals in the justice system. So at this point we’re talking about trying to determine whether these fingerprints are actually unique to individuals and studying – a lot of this research actually occurred in India with English medical doctors and English scientists, and also in South America, where one of the first cases using fingerprint detection was actually heard and prosecuted. We then get to 1892 and we start to see some of these developments actually coming into play, and Galten, who’s one of the fathers of fingerprint detection and fingerprint science, publishes his main textbook. At the same time, he actually convinces Scotland Yard to start using fingerprints as an identification method over body measurements and photos. We then see the prosecution of the first case in Argentina, which involved blood detection, or blood marks, left at the scene. Now, it took a little bit of time for a lot of places to catch on, and in 1902, the UK actually had its first offender prosecuted and jailed by leaving grubby fingerprints at the entry and exit point of a break and enter into a home. Once that then occurred, we started to see a ripple effect, particularly in the UK and Australia, of places picking up fingerprint identification as a method of identifying crooks. So in 1903, you see the first areas in NSW – the first fingerprint bureaus – in NSW and Victoria. 1904, Queensland and South Australia jumped on board, followed by Tasmania, then onto Western Australia, the Northern Territory and then finally the ACT in 1968. So fingerprinting, we were one of the first jurisdictions in the world to actually pick it up. We actually beat the US. Awesome. So when I talk about fingerprint science, we actually combine a whole bunch of different disciplines in order to go through the whole process, from the detection at the crime scene or on a piece of evidence, all the way through to the identification stage. We’re looking at various factors that come into play, from the actual fingermark, or the fingerprint patterns, developing in utero, so we’re looking at physics – the flow of the amniotic fluid actually has an effect on how your fingerprint patterns form over your development. We also see that biology comes into play in this as well. Our metabolism, our biochemistry and what we eat all factor into how our fingermarks actually, or what deposits we have in our latent fingermarks. Physics are also important in how we actually put those fingerprints down, so when we touch an object, whether we slip, whether our skin distorts as we put those fingerprints down, is a factor of physics. And we also have chemistry involved as well with the latent secretions, or our sweat secretions. Chemistry and physics are also really important in the detection aspect. We can’t actually visualise latent fingerprints – latent being hidden or invisible – until we do something to them to manipulate them or make them visible. In this instance, we often use light or a chemical technique. Computer science and coding have given us the programs that then allow us to do the identification stages and to search large databases. Instead of going through 10-print cards and going through every individual ones, we now have systems such as NAFIS, or AFIS, which do the searching. Obviously we still have that human aspect involved – we do need an expert to actually do the comparison process. But we also have statistical methods that can help us, and more and more in recent times, understand the level of uniqueness or variability in certain patterns, how often they occur, and statistcs has even been used to understand how fingerprint ridges and the points of identification that we use actually form during gestation itself. So I want you guys to do a little bit of an experiment with me. Rub your hands together and see how they feel. Do they feel a little bit damp, do they feel really dry, are they a little bit greasy? So take note of how they feel. Now, what I want you to do is rub your hands against your forehead or through your hair. Okay. Now rub your hands together again. Now, any of my students in the audience will know exactly what I’m doing here, but was there a change in how your hands felt after that process? Did they feel a bit greasier, did they feel a bit sort of slicker, or different to that sweaty feeling before? Well, this is actually due to the different secretions that we have over our bodies. So if we take a latent here, so this is actually quite a visible fingermark; it’s from a really, really good donor. Basically you put your finger down onto a surface. If we zoom in on that a bit, we can see that there’s actually a mixture of various different components there. Now, when we take that down to the molecular level, we start to see thousands and thousands of different compounds. Everything you eat is actually secreted in some way, and it pretty much appears in your fingerprints. So we get the sweaty component, or the perspiration, is full of amino acids, peptides, sugars, metal ions, and even things like drug metabolites, such as nicotine or even caffeine metabolites as well. So that coffee you may have had this afternoon, or that tea, is probably starting to come through in your fingerprints about now. We also get that more greasy component that you felt by rubbing your hair or your forehead, is actually the sebum, and it’s full of lipids and other materials and fatty acids. We even get cholesterol and other hormones in there as well. Now, of course, not all latent fingerprints that are left at a crime scene are that fantastic. We can’t just go up, take a photo and walk out. A lot of them are actually not that easy to detect, or are completely invisible to the naked eye. So the big questions, and the big challenge, within fingermark detection, is how do we visualise the fingerprint pattern without actually destroying it. We’re not doing chemistry in the sense that we can extract the material and then run it through a high resolution or high sensitivity instrument; we actually need to keep the fingerprint there in its entirety in order to do the identification step later. So most of you would be pretty familiar with fingerprint powders. Since fingerprints have started to be used for detection processes and identification processes in forensic science, powders have pretty much been the mainstay, and the longest serving technique. We then see the first chemical technique, ninhydrin, which was actually used to stain TLC plates. And that popped up in the mid-1950s. It was then in the late 70s through to the mid 80s that we start to see the golden age of latent fingerprint detection, and we’re seeing a lot of physico-chemical techniques, so a physical and a chemical aspect to the technique, such as vacuum metal deposition, superglue fuming, and metal deposition techniques, such as physical developer and multi-metal deposition. The wheels didn’t slow down at that point. In the early 90s through to the mid 2000s, we start to see the rise of the next generation of amino acid reagents, so the next generation of ninhydrin-type compounds. And we also see a new and improved version of MMD, which is single metal deposition. And then we get to the late 2000s, so we’re in this area around about 2007 through to now, where we see this absolute explosion in detection techniques. We’re seeing chemical imaging pop up, we’re seeing biorecognition techniques, and nanoparticle and lipid stains. Now, you’d think that we’d pretty much solved the conundrum of finding every single latent fingerprint that’s there to be found. We’ve actually hit a plateau. We’re finding that the more proposed techniques that come out, we’re only seeing incremental improvements in that detection level. So we’ve actually hit this point where we’re producing these great ideas from new science, but we’re not getting any closer to finding those really weak, or really hard to detect, marks. Now, we don’t know why – or we’re trying to get to the bottom of why this is. And at the moment, it occasionally feels a little bit like this – we’re running around in circles. However, there is a way around this. We do need to start looking at knowing what we’re actually dealing with, and this is where part of our research is actually really taking off at the moment. So it’s one thing to say that we can build more and more sensitive detection techniques by looking at new chemistry and new instrumental techniques, but we really need to focus on what the fingerprint is actually on, because that’s a massive part of the equation. So if we understand the interaction between a fingerprint and the surface that it’s placed upon, and all of the factors that actually influence that process and how a fingerprint ages over time, how it weathers and how long it can stay on that surface, we can actually get an indication of why we’re hitting that plateau, and how we can get around it so we can build more useful and better techniques. So for example how different types of plastic actually [inaudible] a mark over time, or how different types of paper interact with a fingerprint reagent, and may give us really good results or really, really bad results, and how that actually changes with the technique that we choose. We’re also building a better understanding of how some of the older techniques, such as physical developer, actually work. So the assumed process for some of these techniques that has been around for decades is not actually the case – we’re learning new science every day, and it really takes a plucky young PhD student to challenge things and to just keep asking questions. Those of you who are prospective young scientists in the audience, never stop asking why. It may drive everyone around you nuts, but it gets stuff done. So once we’ve actually built an understanding around these processes, we need to start looking at how we can locate or detect our latent fingerprints. Some of the other research that our students are looking at is moving processes, whether they should be done at the crime scene or at the laboratory. Is there a way, for example, that we can lift and go, or grab and go, a fingerprint, and then detect it later? So can we just get in, get out, without actually having to pull out the powder brush or the superglue wand and do the detection process at the scene?

Also, looking at whether DNA and latent fingerprints can be either detected together, or analysed in a sequence. Now, as our DNA detection methods and our DNA profiling methods get more and more sensitive, we can pick up smaller and smaller amounts of DNA – potentially what we even see within a fingerprint. And this is a big question that’s being asked by a few research groups around the world at the moment, and it’s certainly something we do see popping up in case work as well. Coming up with better techniques to develop marks and also better sequence on how to actually detect those marks in the first place. Now, unfortunately we don’t have one universal, golden technique that works on absolutely everything. WE need to use a sequence of techniques, or a combination, in order to get the best possible chance of detecting a latent fingerprint. Now, in these instances, we do have fairly established sequences for some processes, but then not for others. For example some of the work that we’ll be doing at AFTER looks at the best processes for not only cleaning up evidence that has been around a decomposing victim, but then looking at what the best sequences are in order to actually detect any fingerprints on that evidence. At the moment, there have been no studies into this, and we actually don’t know what the best process is – there’s nothing written in any of the procedural manuals. We can hazard a guess, and we can work off our best understanding, but it would be good to have the data with us. And then there’s also the question of how many fingerprints can we actually detect? At the moment we don’t know hard numbers as to how effective some of these techniques are. We know that they work; we know that they work a lot of the time, but we don’t know and we don’t understand what we’re missing and why we’re missing it. And again, this is a really important aspect in trying to come up with better detection methods, and really propel fingerprint detection into the next decade and beyond that s well. We’ve already been around for about 110 years; we’d really like to stick around for another 110. And of course, the overall idea behind this is just finding that print and being able to do the comparison stage, or the identification stage, later. The final aspect is looking at some of the new techniques we can actually look at, so Claude mention bio recognition techniques, and this is a big part of what my group is also studying. So can we use biomolecules in order to detect those marks? We know that biomolecules tend to be quite selective, and they can also be very, very good at detecting small amounts of material. How do we exploit those processes in order to get a fingerprint that may not be detected by standard techniques? How do we then use that to exploit other information or other types of marks that aren’t developed at the moment, or don’t have a specific detection process for them? Also looking at some new technologies and new things such as metal organic frameworks, and other luminescent techniques, which are quite novel and quite new within the scientific disciplines, and whether they can actually be used again for getting that better fingermark detection. So now that we’ve actually detected the fingerprint, we’ve got an image there that we can work with, we need to move onto the next step. At the end of the day, the end goal is actually doing that comparison and that identification process. So how do we actually go about this? The main process that’s used at the moment is known as [inaudible]. So the online, or the database such as AFIS will throw out a list of possible candidates, and then it’s up to the practitioner to look through those and actually do the analysis, comparison and evaluation on their own terms. So this part is actually where the human factor comes in. So these are actually fantastic examples – you usually wouldn’t see this level of clarity in the latent on the left. Now, they mark up all of the areas: they assess the quality of the mark, they assess whether it’s useful for identification, and then they start marking up the minutiae, or those points of identification or comparison. Once they’ve done that, they then look at the candidate reference print, and then do the comparison and then make a decision based on any differences between those marks and the similarities between those marks. Now, some of the issues that we have arising from this process are not from the actual process itself; it’s from the quality of the latent print. So, a lot of you have probably heard challenges that have come up to latent fingerprint identification in the media – it’s been quite a big thing since the National Academy of Science’s report in 2009, and we see these stories pop up in the media quite frequently. They just kind of pop up, disappear, be quiet for a few months and then pop up again. So in the instance of cases such as the Madrid bombings, the marks are actually quite distorted or there’s quite a poor quality latent. Now, there were a lot of other issues with that case, but one of the confounding factors was actually that the latent print was really poor quality. So one of the issues that we have to think about is how do we detect whether there is distortion, how do we quantify how much distortion is there, or a level of poor development or a missing area of detail, but then how do we determine whether that’s actually two prints looking similar or different is due to those distortions, or whether it’s actually due to something that we call a close non-match? Take for example these two coins here. Look for similarities between these two coins. They’re both gold coins, they’re roughly the same size, they’re roughly the same thickness, and both are actually of the same value. They also have a figurehead on the head side of a monarch. Now, they look – if you look at the similarities, they actually look pretty similar, but there’s one major difference between them, which is the actual figurehead on the coins: one is a king, the King of Sweden, and one is the Queen of England and Australia. So we actually joked about this at a conference a few years ago as a close non-match in currency. And this is where statistical models and probabilistic models can actually come in, and we’ve got a couple of PhD students who are working on this at the moment. So for example, in the instance of a close non-match, looking at how different those clusters of fingerprint, those minutiae are between a close non-match versus a distorted fingerprint, and also measuring that level of distortion as well, so think about when you put down a latent fingerprint, you press your skin, or your finger, against a surface. If you twist that to a certain degree, you’ll end up with a level of distortion as your skin moulds and moves during that motion – it can be torsional, so twisting, or it can be slipping. Now those o you that are quite young, the younger you are, the more elasticity in your skin, the greater that level of distortion before your finger slips off the surface and moves. So this research is looking at different models, both a more practical model that can be slotted in and give an indicator or a red flag if there might be an issue that it’s not actually a distortion effect, but if it’s a fingerprint form a different person, versus a true distortion that can then be looked at and assessed by the practitioner. Then the other issue that we look at is will fingerprint identification actually get to the point that it’s presented in court as a likelihood ratio or a random match probability in the same way as DNA. And this one is a big argument that’s going on at the moment within the fingerprint community. Now, DNA had its own argument – the DNA wars. They basically came up with – they went from doing an identification statement to a random match probability. Fingerprints may go the same way, but it’s not going to be without a lot of infighting and a lot of resistance. So keeping that in mind, the future is pretty bright, and basically as we move through, we’re looking at this rapid development of new technologies. Now, we’re quite proud here to be at the forefront of this, but we do work with other groups around the world who are also pushing each other along. So when we talk about integrated processing, such as looking at different traces together along with the fingerprints, so chemical residues or DNA, more to the chemical residues than the DNA side, we’re probably not that far away from seeing some of these processes implemented into the forensic science world. And things such as intelligence, so not looking at a straight, 1:1 comparison, but trying to draw links between multiple different places using latent fingerprints as well. Smart fingerprint detection, such as bio recognition techniques, we’ve got our work cut out for us, but I think we’ll actually get something that hopefully is starting to roll out into practice in the next five years. Technology is actually moving at a much more rapid pace than we can really keep up with in this space – it’s just making it work for us that’s the challenge. Building a fundamental understanding of the fingermark’s interaction with surfaces and how all of that physics and chemistry actually affect our detection methods is actually a much larger question than we ever thought it would be – I’m finding this the hard way. So given us a generous 5-10 years to try and sort some of that out. Chemical imaging, so actually being able to take a technique which measures, say for example, infrared spectroscopy or other types of really focused chemical spectroscopy techniques that can identify particular bonds or compounds and give an image of that fingerprint and also tell you what is there. At the moment this isn’t feasible because all the instruments require really, really tiny samples and also require excessively long times to actually do the imaging. That said, technology is on an upward trend, and we’re seeing more and more miniaturisation. Back 5, 10 years ago, we probably wouldn’t have been taking a ramen to the scene, and now you can get them about that big. As for the statistical interpretation aspect, whenever it sorts itself out. Some of the optimists in the field say within the next couple of years – it just depends on how much resistance is actually out there to it. I know a lot of people would like to see us move to a more likelihood ratio model. So last of all, thank you for coming tonight, and thank you to all of these people. Obviously without our research students, a lot of this wouldn’t actually happen, so since I’ve got you guys as a captive audience right here right now, I’d like to say on behalf of our research group, thank you very much.

[Applause]