The study using AI comes from Stanford University and Colorado State and appears in the January 30, 2023 edition of the journal PNAS, the Proceedings of the National Academy of Sciences for the United States.
The AI used is a neural network that has been trained using climate models and fed historical temperature observations. The neural network’s predictions on historic temperature shifts have been dead on which has provided the researchers with a sense of assurance that the AI’s future predictions are likely dead on.
What is the AI predicting? A forecast that has 1.5 Celsius being reached between 2033 and 2035. mean warming of the atmosphere will happen. The study concludes that “even if the climate forcing pathway is substantially reduced in the near term” which means regardless of what actions we take to reduce emissions between now and the mid-2030s, atmospheric temperatures will cross the 1.5 Celsius threshold with a “high probability of reaching the 2 Celsius threshold by mid-century.” A rise of 2 Celsius is equivalent to 3.6 Fahrenheit degrees hotter. The probability of reaching that threshold is weighted as four-in-five by 2060.
So how do you train a neural network to become a climatologist? The research team fed maps of annual mean temperatures including anomalies to it. The maps covered a geographic range of 72 degrees of latitude and 144 of longitude. The neural network then weighed the data input and minimized any perceived biases. Once trained on submitted data the neural network was asked to make sample predictions of temperature rise within a historic frame of reference outside the geographic parameters of the data input. These out-of-sample predictions validated the neural network’s capability of forecasting future climate change.
Noah Diffenbaugh, a Stanford climate scientist, and Elizabeth Barnes, an atmospheric scientist at Colorado State University, affirmed their belief that the world will hit the 2 Celsius rise even as emissions decline over the next several decades. Why? Because of the latency effect of carbon emissions already in the atmosphere. In other words, there already is a 2 Celsius rise built in that will play out by the time in question.
These findings contradict the consensus produced by scientists and policymakers who produce the Intergovernmental Panel on Climate Change (IPCC) reports. Their most recent predictions state that the 2 Celsius mark will not be reached if greenhouse gas (GHG) emissions can be drawn down to net zero before 2080.
Where is the thermometer now? The researchers used an acid test to see if the AI could predict that number. It did this by looking at the history studied between 1980 and 2021 data, stating that global mean atmospheric temperatures would be 1.1 Celsius (2 Fahrenheit) higher than pre-Industrial Revolution temperatures by 2022. That accuracy astonished Diffenbaugh who in a Stanford news release stated that he was surprised by the AI. He states “The fact that the AI has such high accuracy increases my confidence in its predictions of future warming.”
Going forward the AI has predicted atmospheric temperatures to have a one-in-two chance by 2054 of hitting the 2.0 Celsius threshold, and a two-in-three chance of crossing it between 2044 and 2065. The chart below provides a sample of the AI’s low and high prediction levels between now and 2065.
The only way to bend these odds means a concerted effort by all producers of GHGs to reduce the output of carbon dioxide, methane and other atmospheric-warming gasses from every aspect of their operations. It is no longer a question of meeting the lower 1.5 Celsius objective, but avoiding the 2 Celsius scenario with unintended and unpredictable consequences for the planet’s ecosystems.
]]>Stanford University researchers have tackled the challenge to bioengineer plants to face droughts, extreme heat, and floods and their impact on soil and moisture conditions. The goal is to make plants smarter by designing synthetic genetic circuits that can then be introduced into plants to control cell behaviour. The changes allow the plants to adjust to environmental changes, principally, through the modification of root structures.
Jennifer Brophy, Assistant Professor, in the Department of Bioengineering, is one of a team of seven at Stanford who recently published their research in the journal, Science. In an article that appeared in Stanford News, it describes the work and quotes Brophy stating, “Our synthetic genetic circuits are going to allow us to build very specific root systems or very specific leaf structures to see what is optimal for the challenging environmental conditions that we know are coming…We’re making the engineering of plants much more precise.”
The reseacrh has involved the building of synthetic DNA that works like binary code, (computer programs) to instruct plant cells to express specific genes in areas of the plant. Targeting roots is seen as the best way to have plants tackle climate change. Roots that are deep are better at drawing up water and fixing nitrogen. Shallower branching roots are better at drawing minerals that lie closer to the surface of the soil such as phosphorus.
The end result are smart plant logic circuits that program root systems based on what is sensed in soil conditions. So far the team has built more than 1,000 circuits that can manipulate gene expression. In one experiment they imbued glow-in-the-dark traits in the leaves of tobacco plants by creating 188 circuit designs that all worked when loaded into synthetic DNA. A second test was done on , Arabidopsis thaliana, a flowering plant commonly called thale or mouse-eared cress. With it, the object was to insert new code to increase root density and branching.
The plan, now, is to introduce synthetic genetic circuits to commercial crops. The first plant chosen is sorghum which produces biofuels. The goal is to adjust the genetic circuitry of the plant to help it absorb water and perform photosynthesis more efficiently.
Why not use alternate methods like selective breeding to achieve the same results? Brophy notes there is a sense of urgency to the work the team is doing because of current climate change trends. She states:
“Climate change is altering the agricultural conditions in which we grow the plants that we rely on for food, fuels, fibres, and raw materials for medicines…If we’re not able to produce those plants at scale, we’re going to be facing a lot of problems. This work is to help ensure that we will have plant varieties that we can grow, even if the environmental conditions that we’re growing them in become less favourable.”
Stanford isn’t alone in researching ways to tweak plant genetics. Cornell, Boston University, Penn, Oregon and San Jose State, and Lincoln University, to name a few, are hot in pursuit of similar goals. But how does the Stanford research differ from others exploring genetically modified foods and plants? The targeting of the reprogramming by Stanford’s researchers is far more precise to impact only specific cells and areas within the plant where the desired outcome is needed. This means there is no genetic alteration to what the plants are producing for food or other uses.
]]>So how can you suddenly have one powered by a chip that is so tiny that hundreds could fit on a fingertip (see image below) and that can be incorporated into a different applications and uses in form factors no bigger than a shoebox?
The lead author of the research done at Stanford University, Neil Sapra, describes particle accelerators as much more than technology designed to unlock the mysteries of the Universe. He states, “Particle accelerators are commonly used for medical (cancer treatments, imaging techniques) and industrial (ion implantation, sterilization, imaging) applications, but compact accelerators for these applications still take up entire rooms or the backs of semi-trucks. Part of the reason for this is that they are based on radio-frequency waves, which have a long wavelength.”
The particle accelerators he is talking about come in a variety of sizes. The technology has helped to make magnetic resonance imaging (MRI) possible. Particle accelerators have been fundamental to the development of nuclear power stations. Particle accelerators underly the x-ray technology used to boost security in airports. They are useful technologies for radiotherapy and x-ray devices. They are used to breakdown wastewater and smokestack gases to remove contaminants. They help physicists identify new materials, biologists discover new viruses and proteins, and pharmacologists develop new drugs and vaccines. They are even used to create shrink-wrapped packaging.
The Stanford invention is a dielectric laser accelerator (DLA) that uses smaller wavelengths than the CERN. This allows for a significant size reduction by a factor of 10,000. What gets lost in reducing the particle accelerator’s size is velocity. Big accelerators can ramp up to near light speed as they endeavour to smash atoms into subatomic components. But those kinds of speeds aren’t needed for industrial applications. States Sapra, “The ultimate goal … is to achieve an accelerator capable of producing 1 MeV [one megavolt] energy beams which can be a truly compact (imagine shoe-box size or even smaller) particle accelerator for medical, industrial, and scientific applications that can be widely and cheaply available.”
In the future small particle accelerators like the DLA could become a common household technology. Is this farfetched?
The discoverer of microwaves back in 1864, James Maxwell, (made famous for his equations), would never have contemplated its use in cooking. And yet today, there are very few homes without microwave ovens. In fact, some homes today no longer come with conventional ovens.
So Imagine a DLA in the future being used to enhance communications, broadcasting, and in a more distant future teleportation or even time travel. Star Trek, here we come.
]]>
States Julie Parsonnet, School of Medicine, Stanford University, “Our temperature’s not what people think it is. What everybody grew up learning, which is that our normal temperature is 98.6 [Fahrenheit], is wrong.” The study sampled 200,000 using data from a number of U.S. sources.
A recent British study of 25,000 concluded that our average body temperature today is 36.6 Celsius (97.9 Fahrenheit). The American research noted that the temperature of males born in the 21st century averaged 0.58 Celsius (1.06 Fahrenheit) lower than men born at the beginning of the 19th century. The difference for women was less at 0.32 Celsius (0.58 Fahrenheit).
A German physician, Carl Reinhold August Wunderlich, was the first to document normal body temperature ranging from 36.2 to 37.5 Celsius (97.2 to 99.5 Fahrenheit). The median was calculated to be 37 Celsius (98.6 Fahrenheit) and that’s what we learned in school. But in fact, some of us run hotter and some run colder.
When you consider normal body temperature has a variance of 1.3 Celsius (2.3 Fahrenheit) it would suggest that the data collected fits easily within that range.
The Stanford study is an interesting piece of work. The analysis was done on 677,423 human body temperature measurements covering populations spanning 157 years. Could the differences in the 19th century be a reflection of less precise thermometers and methods in obtaining body temperatures, or are these findings real?
The proof in the pudding is the measuring of modern populations because control of the calibration can be readily verified as accurate although biases could exist based on the type of thermometer used and from where the measurement was taken (oral versus other means).
Having taken potential variables into consideration for the more recently collected data, the researchers at Stanford concluded “the observed drop in temperature reflects physiologic differences rather than measurement bias.”
One of the reasons posited is Americans over the past two centuries have increased in size and weight. They are bigger and that might affect body temperature. The other possibility is that metabolic rates have declined since the beginning of the 20th century based on other experimental data.
Besides a resting metabolic rate, decreasing levels of inflammation in the general population because of improved living standards, better heating and air conditioning in homes, fewer chronic infections, less dental disease, and antibiotics could all play a factor. It is particularly interesting to link how much time Americans spend in heated homes in the winter, and air-conditioned homes in the summer in the latter part of the 20th and now the 21st century, a phenomenon described as thermoneutral, and its impact on metabolic response and subsequently body temperature.
The Stanford study summary states, “normal body temperature is assumed by many, including a great preponderance of physicians, to be 37°C …. Our investigation indicates that humans in high-income countries have changed physiologically over the last 200 birth years with a mean body temperature 1.6% lower than in the pre-industrial era. The role that this physiologic evolution plays in human anthropometrics and longevity is unknown.”
]]>
We are not there yet, but already artificial intelligence (AI) is helping doctors to help patients. It’s only a question of time before we can see the middleman moved out of the equation.
Today AI in health care is being applied to improve operational efficiency and to decrease the case burden on physicians. This is where machine learning provides distinct advantages whether it be in documenting, charting, billing, or in the case of helping with diagnoses, recognizing patterns that indicate illness much faster and in greater volume than any human can. Pattern recognition means AI tools can be applied to all types of medical practice from nursing to social work, and to community health. Pattern recognition technology can observe a patient’s behaviours better than most health practitioners. It can always be on closing the care gaps that occur when patients are not being observed by humans. It can help patients to stay in homes rather than occupy hospital beds as it provides around the clock data to medical teams without human presence necessary. It even, in the form of robot assistants, provide basic support services and companionship.
Let’s look at some of the most recent reported advances in AI applied to health care.
There’s a trick that most cardiologists wouldn’t claim to be able to do. A good cardiologist might detect heart failure by a quick look at a patient. Does the patient’s colour look right? Are extremities and fingernails looking bluish? Are there signs of puffiness and swelling in the feet and ankles? But before conclusively diagnosing congestive heart failure (CHF) the doctor would do an electrocardiogram (ECG), a chest x-ray, and have the patient wear a 24-hour halter monitor to provide a 5-minute-segmented record of heart function. Imagine, however, if an AI could draw a conclusion of CHF without all of the above and do it from a single heartbeat. That could not only save time but also help in early intervention when CHF is detected, a condition that 5 million Americans and 26 million around the world live with every day.
That’s what researchers at the University of Surrey have developed using a neural network that accurately detects CHF through the analysis of one ECG heartbeat. Dr. Sebastiano Massaro, Associate Professor of Organisational Neuroscience, at Surrey, is one of the researchers responsible for creating the AI. In a university news release, he states, “We trained and tested the CNN [Convolutional Neural Network] model on large publicly available ECG datasets featuring subjects with CHF as well as healthy, non-arrhythmic hearts. Our model delivered 100% accuracy: by checking just one heartbeat we are able to detect whether or not a person has heart failure. Our model is also one of the first known to be able to identify the ECG s morphological features specifically associated with the severity of the condition.”
The benefit of such an efficient diagnosis without tying up lots of medical resources cannot be underestimated.
Breast cancer is the most common cancer suffered by women globally. It is also the second leading cause of death. My wife goes through regular screenings for breast cancer, as do most women in North America and Europe. The tests include mammograms, ultrasound and physical examination. Women are taught to self-examine for early detection. The problem, however, is that screening results are 20% of the time wrong with 50% of women receiving annual mammograms experiencing at least one false cancer positive result within a 10-year period.
In the study reported in Nature, published online on January 1, 2020, Google’s DeepMind AI outperformed six radiologists in the U.S. study in interpreting mammogram results, reducing false cancer positive error rates by better than 350%. At the same time, it helped reduce workloads for those assigned in the United Kingdom to do second readings of mammogram results by 88%, a common practice there.
For the purpose of the study, the AI system provided a comprehensive report, not individual results. But it could be automated to give immediate feedback during screening which would alert radiologists to detected cancers.
A neural network developed at Stanford is being used to detect and diagnose skin cancers using imaging analysis. The researchers trained the convolutional neural network using a dataset of almost 130,000 clinical images and over 2,000 diseases and tested its performance against 21 board-certified dermatologists. The neural network achieved a performance equal to the dermatologists.
The goal of Stanford is to turn the AI detection tool into an application that can be put on a smartphone, capable of taking an image of a suspicious skin appearance and competently detecting carcinomas and malignant melanomas.
Early detection of Alzheimer’s is critical to treatment to stem the progression of the disease. Diagnosing it when clinical symptoms start manifesting themselves often comes when too many of the brain’s neurons have already died making it irreversible.
But an AI machine-learning algorithm trained to analyze a massive dataset of PET scans of patients eventually diagnosed with Alzheimer’s, mild cognitive impairment, and no disorders at all, successfully identified between 92 and 98% of patients who eventually developed the disease. Diagnosis occurred more than six years before doctors confirmed the disease.
The study was first published in Radiology in 2018. Sixteen studies since then have looked at a number of AI deep learning and machine learning algorithms aimed at early diagnosis of Alzheimer’s with accuracies over 98%.
For Alzheimer’s research AI represents a diagnostic breakthrough that may lead to much better treatment and even remission of the disease because of early detection.
Stem cells derived from bone marrow produce the cellular components of blood. Hence they are called blood stem cells. The treatment that is being described at the Orlando meeting is blood stem-cell transplants. What’s involved?
Defective blood-making cells are identified and eradicated. They are then replaced by healthy ones harvested from donors or patients themselves. The transplanted cells take hold and differentiate into the components that make up healthy blood.
How does this differ from current practices?
Today’s medical standard is to irradiate or treat with chemotherapy drugs a person suffering from one of these diseases. This kills off the patient’s existing blood stem cells and a host of other cells in the bone marrow. The side effects of this type of treatment can compromise the immune system, cause infertility, and seed cancers that show up later in the patient’s life.
In an article appearing in Nature on November 29, 2019, author Heidi Ledford uses the analogy of a hotel where the old treatment requires all guests to leave, versus the new one which targets specific guests only.
Two biopharmaceutical companies, Forty Seven and Magenta Therapeutics are doing clinical research targeting the blood stem cell molecule, c-Kit. c-Kit is a protein found on the surface of cells that binds to stem cell factor, responsible for the growth of certain types of blood cells. c-Kit is associated with a particular protein tied to the CD47 gene. Inhibiting this CD47 protein boosts c-Kit’s effectiveness in targeting blood stem cells containing the protein, killing them. This allows transplanted healthy blood stem cells introduced into the bone marrow to thrive and unlike chemotherapy or irradiation does not as a consequence destroy the immune system.
At Stanford University associated with the biopharmaceutical company, Amgen, other research is looking at CD117, another antibody that effects c-Kit. Using donor blood stem cells in a limited clinical trial involving six babies, four were effectively treated for a genetic immune system disorder.
European researchers have also been testing the technique on an immune disorder called Adenosine deaminase deficiency (ADA) a genetic condition that causes Severe Combined Immunodeficiency (SCID). People with this genetic disorder cannot fight off most infections. By genetically modifying removed blood stem cells and reintroducing healthy replacements, the research shows promise in treating the condition. It may also help in tackling diseases like beta-thalassemia, sickle-cell disease, type 1 diabetes, and systemic scleroderma.
What is happening to the body with these blood stem cell treatments? It is believed that introducing healthy blood stem cells in these various procedures is helping to reset the body’s immune system and may soon become a new protocol in dealing with blood and immune disorders.
A downloadable paper describing their discovery, published July 8, 2019, by the American Chemical Society, and entitled “Charge-free Mixing Entropy Battery Enabled by Low-Cost Electrode Materials,” describes the availability of “salinity gradient energy” in estuaries where fresh and seawater mix, estimating the potential “blue energy” yield at 2 Terawatts. It acknowledges that tapping this energy would be challenging, but a feasible starting point would be to use the technology in association with wastewater treatment plants discharging into saline environments. Coastal wastewater treatment plans could, it states be “a significant untapped source of energy” representing 0.65 Kilowatt hours per cubic meter of treated wastewater. The paper proposed that the estimated recoverable energy from wastewater treatment plants globally would amount to 18 Gigawatts. At the least the blue energy yield from a coastal wastewater treatment plant would be more than enough to provide the facility’s energy needs.
It uses the natural electrochemistry that exists between saltwater and freshwater when they interact. The higher salt concentrate in the seawater causes molecules from the freshwater to interact creating pressure that can be turned into electricity.
Called a mixing entropy battery or MEB, it uses low-cost, commonly available battery electrode materials to convert salinity gradient energy into high-voltage electricity in a four-step process:
Recoverable energy from wastewater effluent is rated at 89.5%, and from seawater, 97.6%. Capacity retention after 50 to 150 cycles of operation, based on the electrode materials chosen varied from 93 to 99%.
States Kristian Dubrawski, a post-doctoral scholar in civil and environmental engineering at Stanford, and a co-author of the paper, “Blue energy is an immense and untapped source of renewable energy… Our battery is a major step toward practically capturing that energy without membranes, moving parts or energy input.”
The significance of MEB is the technology can work anywhere fresh and saltwater mix. But using it in coastal wastewater treatment plants as a renewable energy source makes sense because these facilities are energy-intensive drawing up to 3% of the U.S. total grid capacity.
Theoretically, if MEBs were to be deployed a cubic meter of freshwater when mixed with seawater would produce enough energy to power the typical American home for 30 minutes. If brought to scale, the technology’s potential is mind-blowing.
The technology doesn’t require any expensive materials. There are no moving parts. It doesn’t need expensive membranes or instrumentation to control the charge and voltage. The electrode materials cost between $1 and $3 per kilogram. It can be scaled to not only provide the energy for wastewater treatment, but in places like California and Florida, could produce enough surplus energy to power nearby desalination plants.
Dubrawski calls the invention, “a scientifically elegant solution to a complex problem.” And although it doesn’t address blue energy on a global scale, “it is a good starting point that could spur these advances” turning every river that empties into the ocean at its mouth into an enormous power source using MEB technology.
]]>
The study estimates that between 3 and 20% of past intranational conflicts, civil wars, and other internal uprisings, can be attributed to climate in the previous century. One of the co-authors of the paper, Philip Roessler, at William & Mary University, talks about how some have described the civil war in Darfur, Western Sudan, as the first climate-change conflict. But he believes social tensions between ethnic groups combined with a weak national government contributed greater to the violence there versus long-term drought conditions.
Mach describes the role of climate change as having significant social costs which could exacerbate social tensions, damage economies, negatively impact food supply, and contribute to growing inequality to become drivers of conflict. All who contributed research to the final paper agreed that climate is playing a role in organized armed conflicts in recent decades, and that future climate change will prove to be even more disruptive because conditions will alter beyond our human experiences to date.
James Fearson, of Stanford, a co-author, states that “putting nontrivial weight on significant climate effects on conflict is reasonable” in the rest of this century because of anticipated changes to the environment.
The study recommends risk reduction strategies to minimize or mediate conflict including:
Probably the most pressing issues will arise around food, freshwater and energy security as atmospheric temperatures continue to rise. Any destabilization of one or more of these three could become conflict drivers.
What surprised me about the study was the lack of reference to older conflicts where historians suspect that variations in natural climate played a role. Again, and again, there are historical references to drought and other weather phenomena influencing the movement of tribes from Central Asia to the periphery of the empires of antiquity: Rome, Persia, and China.
It also surprised me that the study didn’t look at the relationship of climate change to outbreaks of pandemics and the combined influence of these two yielding increased rates of conflict.
In the last couple of decades, as climatologists have developed more sophisticated climate models, projecting them back as well as into the future, researchers have analyzed historic movements of people and seen past conflicts in a new light recognizing that climate variation has heavily influenced history.
]]>
The study notes that if temperatures can be held to 1.5 Celsius then 71% of the world’s countries representing 90% of the world’s population would have a better than 75% chance of experiencing positive economic benefits including the three largest economies: the United States, China, and Japan. Countries that would gain the most are the poorer ones predominantly located closer to the equator where lowering the warming would lead to a far better outcome in terms of GDP growth. Even small reductions in the global temperature target limit would substantially improve the GDP of these countries. Northern countries, however, would see the least growth in GDP based on the study with higher temperature increases than the mean impacting them negatively.
Using GDP as a measure to calculate climate change impacts based on varying temperature rise is a novel approach to analyzing future trends. The numbers in this study far exceed any in previous forecasts which talk about hundreds of billions of dollars in GDP impacts versus multiples of trillions.
More disturbing is the conclusion drawn by the authors noting that although the Paris Climate Agreement calls for a 1.5 Celsius optimal target, and a 2 Celsius upper limit, when looking at actual Nationally Determined Contributions (NDCs) of the signatories it is far more likely that we will see mean average temperature rise by 2.5 to 3 Celsius (4.5 to 5.4 Fahrenheit) globally by end of century. They further conclude that failing to meet NDC commitments that yield a lower mean average temperature rise will produce substantial reductions in global GDP, in excess of 25%.
There is an additional author caveat. They study notes it cannot account for “historically unprecedented changes” such as large-scale sea level rise from the disappearance of land-based ice in Antarctica and Greenland, and remark that at 2 Celsius or higher the unprecedented would be far more likely to happen than at 1.5 Celsius.
Peer review of the study praised the work but also pointed out its limitations including no way to measure the impact of technological innovation that could alter temperature rise or technologies that could increase energy efficiencies to make adaptation less economically damaging. Nor does the study look at changes in the use of fossil fuels and how that would contribute to improvements in the ecosystem of the overall global community which would increase GDP even more than forecast.
The authors, however, point out that it is in the best interest of all countries to meet the 1.5 Celsius upper limit regardless. It means the quality of life for most of the world’s population will not have to endure significant disruption.
In a journal article appearing in Nature Energy yesterday, I was quite excited to read about a new solar water cooler created by a trio of Stanford researchers: Eli Goldstein, Aaswath Raman, and Shanhui Fan. The device, based on their initial tests, lowers air conditioning electrical demand by as much as 21%.
In the Nature Energy article, entitled, “Sub-ambient non-evaporative fluid cooling with the sky,” they describe the global dilemma we face with cooling demand expanded to increase 10-times between now and 2050 as temperatures climb and Developing World countries install air conditioning.
How are these new water cooling panels constructed and how do they work? Seen in the picture below they consist of three components:
The thin-film plastic reflects 97% of sunlight by converting the energy to infrared which then gets reflected back into space. This creates a radiative sky cooling effect which lowers the temperature of the air inside the sealed housing which in turn lowers the temperature of the water circulating in the copper tubes contained within it.
The researchers tested three of their panels on the rooftop of a building on the Stanford campus, and the paper published yesterday describes their results. The water temperature in the copper tubing dropped by 5 Celsius (9 Fahrenheit) degrees over three days. The radiative cooling worked.
Interestingly, the thin-film plastic material used by the Stanford team is similar to a technology I recently described invented by engineers at the University of Colorado in Boulder.
Applying the data from the three-panel experiment the Stanford team ran a computer simulation using a two-storey office building in Las Vegas as an ideal model. They wanted to calculate the potential savings their radiative sky cooling could produce and Vegas was the best place to consider since its climate is both hot and dry. The simulation showed very promising results. Over several summer months, the building would have saved 14.3 megawatt-hours of electricity with the range of savings varying monthly between 18 and 50%. Average energy reduction over the entire simulation totaled 21%.
Not surprising, the researchers have created SkyCool Systems, a new venture that aims to commercialize the technology and are conducting a first large field trial on building rooftops in Davis, California.
]]>