Tackle the hard problems.” That’s the key to innovation, at least for Reginald Beer, an engineer at Lawrence Livermore National Laboratory. “I don’t like working toward incremental solutions,” he said. “Give me the problems that are currently unsolved, the ones that have a clear need.”
This philosophy has served Beer well in his career at one of the nation’s premier research and development laboratories. At LLNL, he has tackled some extremely challenging problems, including rapid PCR (polymerase chain reaction). “I was sitting in the pediatrician’s office with my daughter several years ago, and it struck me that if we could make PCR fast and easy enough for use in doctors’ offices, it would have a huge impact.”
PCR is an indispensible technique in medical and biological research laboratories around the world. It allows researchers and clinicians to produce millions of copies from a single piece of DNA or RNA for use in genome sequencing, gene analysis, heritable disease diagnosis, paternity testing, forensic identification and the detection of infectious diseases. The standard approach to PCR typically takes about an hour, which is a vast improvement over pre-PCR techniques that required days. However, PCR for point-of-care, emergency-response or widespread monitoring applications needs to be faster still—on the order of a few minutes.
“When I set out to do this, people said it would never work,” Beer recounted. “But I’ve found that when something’s been tried before and didn’t work, it’s often because they didn’t have tools that were good enough at the time or lacked knowledge that wasn’t available then but is now.”
This was the case with rapid PCR. “When I talked with people in the field, they gave me mechanical reasons or enzyme kinetic reasons why it wouldn’t work. But I answered, ‘nobody really knows.’”
Beer set out to find out just how fast PCR could go. His goal was to develop a device where the limiting factor would be enzyme kinetics or thermodynamics, not mechanical considerations. He also wanted to use volumes that are easy to interact with, not microfluidic-scale, knowing that in a diagnostic setting multiple analyses are often needed and the ability to load samples by hand is vital.
As described in a recently published paper in the journal Analyst, Beer and his colleagues created such a device and demonstrated PCR times of less than three minutes. The device achieves its extremely fast thermal cycling through the use of a porous material and a thin-film resistive heater, making possible heating and cooling rates of 45 degrees Celsius per second, for a thermal cycle speed of less than 2.5 seconds. “This device is unique in that it cools as fast as it heats,” Beer said.
With the device in hand, Beer’s next challenge was to see if any commercially available polymerase enzymes could work rapidly enough. Out of a group of 10 polymerases advertised by their manufacturers as compatible with fast thermal cycling, two enzymes worked in the Livermore device right out of the box.
“We were really encouraged by the fact that two off-the-shelf enzymes worked at these speeds,” said Beer, “and there are a lot of parameters that can be adjusted to potentially go even faster.”
The researchers demonstrated their PCR device by amplifying genomic DNA from an Enterobacter bacterium and a portion of SARS DNA. The first tested the device’s ability to rapidly amplify a large DNA segment, and the second showed the device’s utility in handling a public health threat virus. The device achieved 30-cycle (billion-fold) PCR amplification of the target DNA in as little as two minutes and 18 seconds.
Now that Beer and his team have demonstrated sub-three-minute PCR, they are working to develop a real-time-detection device. They envision a PCR instrument that can complete a test, from sample to results, in five to ten minutes.
The market for such a device would be huge. In addition to the traditional public health and medical research applications, an easy-to-use real-time PCR device would be enormously useful in the livestock, poultry, agriculture and processed food industries for ensuring food safety.
Another key to Beer’s success as an innovator is his use of multi-disciplinary research teams. “You need lots of bright people with different perspectives to solve the really hard problems. You need very creative people who are willing to cross boundaries and try crazy ideas.”
Beer also has found his diverse technical background to be enormously valuable. While most of his research has been in microfluidics, he also has academic training and experience in aerospace engineering, structural dynamics, robotics, biomedical technology, radar applications and optical systems.
“Outside of microfluidics, I’m not an expert in any one field,” Beer explained, “but I know a fair amount about a lot of things and so I can talk to the experts in many different fields and serve as the interface to pull everyone together into a multi-disciplinary team. Since I’m not rooted in any one area, I’m something of an outsider and this can be very helpful. As an outsider, I can question the assumptions and challenge the conventional wisdom. I can bring a fresh perspective and look at the problem from a different direction.”
Often what happens is that technologies used in one field can be transformed and applied to new problems. This was the case with his work on the challenge of detecting improvised explosive devices (IEDs).
“IED detection is a really hard problem,” Beer said. “You need a way to detect all sorts of different devices emplaced in all sorts of different settings. Not only do you need a really good detector, but you also need a way to separate background clutter from a signal that you absolutely have to detect.”
Beer and his team took the lab’s successful micropower impulse radar (MIR) technology and developed signal processing tools to image and detect buried objects in real time. The idea was to apply Livermore’s vast experience in high-performance computing to a challenging detection problem. The researchers have since demonstrated a complete system to detect buried objects in real time utilizing a multi-static ground-penetrating radar (GPR) array. The imaging system essentially reconstructs a tomographic image, like a CT scan, of the objects under the array, and automated signal processing algorithms tease the target out of the surrounding clutter. They also developed a mobile supercomputer to perform the necessary calculations “on the fly.”
MIR is an outgrowth of an ultrafast solid-state digitizer designed to measure sub-nanosecond events generated by laser fusion experiments. MIR devices emit short voltage impulses at a rate of one million per second that are reflected off nearby objects and detected by its high-speed sampling receiver. The pulses can penetrate a wide range of materials including rubber, plastic, wood, concrete, glass, ice and mud, and MIR devices are very compact, with both the radar transmitter and receiver contained in a two-inch-square package.
MIR technology has been licensed for a wide variety of applications, such as construction, automotive safety, and industrial automation products. Other potential applications include security, energy conservation, medical diagnosis and monitoring, residential and commercial automation, material and structural evaluation, underground anomaly detection, and even voice recognition and communications. “Now that we have a real-time image processing and object detection capability to pair with our GPR, we are exploring commercial applications that will benefit from this capability,” Beer said.
When asked about his approach to innovating, Beer explained that he takes advantage of a large network of contacts, including operators, systems engineers and subject matter experts. “We bounce ideas off each other and ask: ‘could this possibly work?’ We rarely sit around in brainstorming sessions—we typically work at a whiteboard in someone’s office.”
Beer noted that in order to solve the hard problems, innovators can’t be afraid to fail. In fact, his goal is to fail fast in order to learn quickly if the approach he’s taking isn’t going to work. That way, he can move on to try the next idea.
“When you’re attacking problems that no one’s solved, it isn’t failure so much as trying a series of ideas to find the ones that work. Failure would be to give up on the problem altogether. At Livermore, we don’t do that.”
Lauren de Vore is a writer at Lawrence Livermore National Laboratory.