Artificial intelligence is quickly growing in importance in the ‘smart building’ sector. Paul Skelton looks at the road ahead for a complex technology.
When Mark Chung received an unexpectedly high $500 monthly electricity bill, he turned to his utility for help and answers.
However, despite ‘smart’ meters being installed in his home, they were no help. So Mark – an electrical engineer trained at Stanford University – took matters into his own hands. He bought some inexpensive meters from his local hardware store, converted them to be WiFi-enabled, then created an electrical map of his home. This showed that a small failure in his pool pump was causing a huge current overload, which couldn’t have been detected by older methods. More important, he learned how hard it is to get information from buildings, which typically lack any kind of computerised management.
Thus was born the idea for Verdigris, a start-up aimed at conserving energy in buildings using artificial intelligence (AI) powered by a graphics processing unit (GPU). Buildings consume about 70% of the world’s electricity and waste 60% of the consumption. That’s $100 billion wasted on electricity each year. To tackle the issue, Verdigris developed proprietary wireless sensors that are clamped onto electrical mains, panels and circuits. The Verdigris digital system then uploads electricity consumption data to the cloud, 24/7. From there, it can sell the raw data to building managers or apply its own AI algorithms and provide any insights it gleans. It can even integrate the data with building management systems to automate electricity usage controls.
By forecasting problems and determining areas for optimisation and automation, Mark, who serves as chief executive of Verdigris, says the company can help facilities to get more done. Hotel managers, for example, could detect and fix building issues before guests even notice them. All this wouldn’t be possible without the plummeting cost of bandwidth, sensors, data collection, processing and storage – which has fuelled the growth in AI.
“AI is extending into every facet of our lives: how we travel, how we produce food, how we work, how we live,” Mark says.
“Smart buildings o..er valuable and extensive opportunities in this trend.”
ARTIFICIAL, NOT FAKE
Long the domain of science fiction, AI is the subject of a common TV trope whereby a computer or robot becomes sentient and learns the power of human emotions (mainly love). In reality, AI isn’t about developing childlike androids voiced by Haley Joel Osment. Instead, it’s about effectively using technologies to simplify our home and work lives. And it’s not just something that sits on the periphery of the market.
In September 2016, Amazon, Google, Facebook, IBM and Microsoft announced the creation of a non-profit organisation aimed at advancing public understanding of AI and formulating best practice for the challenges and opportunities in the field. Further, NVIDIA – a company perhaps best known as a manufacturer of GPUs for gaming consoles – has shifted its focus to the world of AI. It offers organisations the capability to do research into how AI can monitor traffic patterns, treat heart disease and allow computers to interpret sign language.
(NVIDIA is primed to be perhaps the most significant company in the systems integrate on space, so pay attention to everything it does.)
Most recently, the company has announced an arrangement with video surveillance specialist Hikvision to implement AI technology in security cameras. Hikvision Oceania general manager Daniel Huang says that in the past two years, deep learning technology has excelled in speech recognition, computer vision, voice translation and much more.
“It has even surpassed human capabilities in the areas of facial verification and image classification. Hence, it has been highly regarded in the field of video surveillance for the security industry.”
Daniel says the rise of ‘deep learning’ has had a profound influence in the application of intelligent video in target detection, tracking and recognition.
“When applied to those three functions, deep learning potentially touches upon every aspect of the security video surveillance industry.
“It involves facial detection, vehicle detection, non-motor vehicle detection, facial recognition, vehicle brand recognition, pedestrian detection, human body feature detection, abnormal facial detection, crowd behaviour analysis, multiple target tracking, and so on.
“All of this starts to change the focus of security from being reactive to being able to predict problems.”
Hikvision has taken AI technology and developed a family of products to maximise its use. The DeepInview IP camera range and the DeepInmind NVR (network video recorder) range that Hikvision will soon introduce, work together to provide all the power and benefits of deep learning. Cameras provide the ‘eyes’ of the system, and the NVR represents the analyti c and storage capabili.. es of the brain. The products help to tackle security on two fronts – recogni.. on, monitoring and coun.. ng of people, and recogni.. on and detec.. on of vehicles.
“The inspiration for deep learning comes from the human brain’s neural networks. Our brains can be seen as a very complex deep learning model. Brain neural networks are made up of billions of connected neurons, and deep learning simulates this structure.
“These multi-layer networks can collect information and perform corresponding actions. They also possess the ability for object abstraction and recreation.
“In total, there are three main reasons why deep learning became popular only in recent years and not earlier: the scale of data involved, computing power and network architecture.”
Daniel says improvements in data-driven algorithmic performance have accelerated deep learning in various intelligent applications over a relatively short time.
“Specifically, with the increase in data scale, algorithmic performance improved as well. Accordingly, user experience has improved and more users are involved, further facilitating a larger scale of data.
“Video surveillance data makes up 60% of big data, and that amount is rising at 20% annually. The speed and scale of this achievement is due to the popularisation of high definition video surveillance. HD 1080p is becoming more common and 4K and higher resolutions are gradually being applied in many important applications.
“Further, high-performance hardware platforms enable higher computational power. The deep learning model requires a large amount of samples, making a large amount of calculations inevitable.
“In the past, hardware devices were incapable of processing complex deep learning models with more than a hundred layers. In 2011, Google’s DeepMind used 1,000 devices with 16,000 CPUs to simulate a neural network with about one billion neurons.
“Today, only a few GPUs are required to achieve the same sort of computational power with even faster iteration. The rapid development of GPUs, supercomputers, cloud computing and other high-performance hardware platforms has allowed deep learning to become possible.
“Finally, the network architecture plays its own role in advancing deep learning. Through constant optimisation of deep learning algorithms, better target-object recognition can be achieved.
“For more complex applications such as facial recognition – or in scenarios with different lighting, angles, postures, expressions, accessories, resolutions, etc – network architecture will affect the accuracy of recognition. That is, the more layers in deep learning algorithms the be.. er the performance.”
Given that the new Hikvision cameras are not yet available, and that many of the other developments in the world of AI are largely hypothetical, integrators could be forgiven for dismissing the technology as something way in the future. However, it seems they would be wrong. In 2016, the USbased start-up Josh.ai introduced to market a control platform built on the concepts that form artificial intelligence.
Josh is a voice-controlled automation system that uses natural language commands to direct the smart devices in a home. Further, Josh can be taught with ease. Even complicated queries can be effortlessly programmed, such as: “At sunrise if I’m home slowly fade the bedroom lights on, open the drapes, turn on the radio and brew a pot of coffee.”
Although some purists might not consider this to be AI, it is definitely the most significant step towards the realisation of AI in the home. Alex Capecelatro, co-founder and chief executive of Josh.ai, says lots of people have a different definition of AI.
“However, I like to think of it as when you move from home control to home automation.
“Right now, people still flip light switches – this is control. They could push a button on their smart phone, but that’s also control. They could even use their voice to turn on the lights, but again that’s control.
“What we’re really interested in is the point at which you go from control to automation. “This is when you walk into your home and you don’t have to hit a light switch or tell the house what to do. The house knows what to do because of what it has learned.”
Alex says that in many ways this is the AI tipping point and the ultimate goal for Josh.
“When I started Josh, the idea was to use AI from the perspective of natural language (ie: voice control) as well as the ability for the system to learn patterns so that users don’t have to spell out everything. The system can start to become predictive.”
Schneider Electric director of smart space Ben Green agrees that smart devices are just devices until they start doing something more effectively and efficiently. “Voice control is one part of AI in an otherwise larger connected world of artificial intelligence,” Ben says. “Netflix and Spotify run their own algorithms to understand what you want to watch or listen to. When you get into your car, maps let you know about road conditions and tell you how long it will be to your next appointment. “Then you’ve got lighting, blinds and climate control to optimise the home and work environments. “So, AI is already happening, but it’s about how you then aggregate and control all of those layers to create a new world of meaningful outcomes.” Ben says that to understand the full potential of AI integrators must first get to grips with why the technology is becoming more important and why it’s happening now. “Look at the connected world of technology. Predictions from Cisco put us at 50 billion devices by 2020, and we’re all seeing the trajectory of that. But as the number of connected technologies increase, we’re seeing an enormous increase in data analytics. “Big data as an industry, by 2020, is going to double in value. It’s going to become a $203 billion industry. “Now, consider the energy crisis. In the next four years, energy consumption could increase by 50% which will see pricing also increase. It’s understandable then that promoting efficiency is going to be 100% necessary. “AI will be needed to centralise and/ or aggregate these connected devices to create new outcomes. In fact, AI is the only way we’ll be able to navigate that much data.”
The post Artificial Intelligence: Science fiction to science fact appeared first on Connected Magazine.