Smart infrastructure helps make cities more efficient and safer. And it’s making greater strides with collaboration technology.
Today, collaboration tools have joined forces with other technologies to change all aspects of daily life.
Tools like mobile devices, connected sensors and location-based services have converged to bring new benefits to people at work, at home and in the cities where they live.
In this e-book, we look at how these technologies are converging to make service more personalized for consumers and workers more productive at work. This chapter examines several technologies and considers their role in medical care, smart city living and other scenarios.
The problem of exploding city growth has reached critical mass. By 2050, the United Nations estimates that nearly 70% of the world’s population will reside in metropolitan areas. This massive growth strains city services such as trash pickup, water sanitation and public transportation. City growth also puts pressure on the safety and security of residents.
Strained cities have turned to smart city infrastructure to address the challenges of exploding populations. Smart city technology can track a bus is on its route. It can provide city workers with the location of trash to be picked up, potholes to be fixed or even traffic accidents. With sensors that collect environmental information, Internet of Things-connected devices can convey useful information to consumers and services providers.
But these smart sensors remain dumb without the aid of other tools and capabilities. Sensors can gather data on a train’s estimated time of arrival, but collaboration technology provides that data meaningfully, in the proper context and in digestible ways.
So now let’s consider the various technology elements that, when combined smart city infrastructure and collaboration tools, have impact.
Network bandwidth involves both access to a network connection and available bandwidth through that connection. Historically, access to a network was largely limited to fixed network connections or limited footprint Wi-Fi connections. This required individuals to stay close to home or work to maintain that connection. The Pew Research Center estimates that between 2000 and 2016, Internet usage by U.S. adults climbed from 52% to 88%. Over the same period, Pew’s research shows that broadband use increased from 1% in 2000 to 73% in 2016.
For the past two years, fixed network bandwidth in the United States has increased at roughly 30% in total.
The next generation 5G mobile network promises to provide even greater network throughput, reduced latency and improved energy efficiency. As the number of devices grows worldwide—estimated to increase to more than 50 billion by 2020—increasing bandwidth is critical.
In particular, small “smart” devices have little native compute and/or storage are integral to building a smart infrastructure. They have the ability to communicate short data bursts that can be processed centrally. Further, the additional network capacity provides greater latitude in communication payload; devices won’t be limited to relatively simple, single data point transmissions. Instead, these devices could transmit (and receive) richer data transfers. In total, high-speed and readily available network connections enable a highly distributed device population that can connect to an array of centralized compute for data processing.
Additionally, new networking strategies such as edge computing have a profound effect on enabling collaboration technology and smart infrastructure to work together, which we discuss in detail later in this article.
Boosting network bandwidth will be key to providing the connectivity we need to bring collaboration tools and smart infrastructure together. Both technologies are bandwidth hogs and need always-on connectivity to reach their full potential.
From 1956 to 2015, there was a 1 trillion-fold increase in compute performance. Processors in common devices— phones, tablets, video-game consoles—are capable of complex computing operations, from virtual reality environments to near-instantaneous facial recognition.
Simultaneously, the same phenomenal processing advancements have been incorporated into a new, massively scalable cloud-based infrastructure. Amazon and Google have built hyperscale cloud compute infrastructure that provides organizations with unprecedented, centralized computing capability combined with pay-as-you-go economic models. As a result, enterprises can access supercomputer-like power without the cost or maintenance burdens. It’s no longer necessary to buy, manage, maintain or host servers.
In turn, the pace of product development has increased rapidly. New companies avoid the cost and overhead associated with data centers. If an enterprise wants to write a new mobile application, the server infrastructure can be provisioned in minutes—and more cost-effectively with multi-tenant, on-demand cloud models. Developers can write applications to incorporate an array of services into an application. When combined with personal compute devices and ultra-fast network connections, additional possibilities, such as distributed computing models, begin to form.
Over the past 30 years, data storage has gotten dramatically cheaper. As of 2016, consumers can buy multi-terabyte hard drives for about 30 cents a gigabyte. Cloud providers such as Microsoft and Google now provide storage that ranges from tens of gigabytes to a terabyte for free. From 1980 to 2016, data storage costs plummeted per magnetic gigabyte from $437,500 to about $0.019.
“Boosting network bandwidth will be key to providing the connectivity we need to bring collaboration tools and smart infrastructure together.”
One immediate opportunity is with data collection. Storage costs once encouraged organizations to make smart choices about how much of that data to keep. Now that they can store data so cheaply, organizations are apt to do so with relative abandon. This can produce valuable insights – at least for those equipped to wade through the volume of data.
But the goal is for companies to decide what data they need to keep—and that they can fully protect—to generate business insights without storing data ad nauseam. While access to storage is far cheaper, the consequences for breached data have become only costlier. So companies need to balance their need to store, manage and access data with the costs of failing to protect it.
Machine learning and artificial intelligence are also factors. Machine learning applies algorithms to large data sets and uses statistical analysis to provide insight.
Algorithms can be categorized into supervised and unsupervised. Supervised algorithms require humans to train the algorithm across a sample data set and to provide the input as well as an expected outcome. Conversely, unsupervised algorithms require no training and use deep learning to arrive at conclusions not predicted by human operators.
Historically, this kind of technology was accessible only to advanced organizations, government agencies or higher-learning institutions. Machine learning technology is now broadly available and democratized. Cloud providers such as Amazon, Google and Microsoft have made it relatively trivial to provision machine learning capability. This encourages experimentation with, if not outright integration of, machine learning in the enterprise. As a result, new data analysis possibilities are possible for companies of all sizes and budgets.
From Facebook to Microsoft Outlook, common applications now use machine learning algorithms to enhance user experience, make tasks easier and connect ideas. These common uses help to shape Twitter news feeds, identify spam email, reduce background noise from a speaker’s voice and automatically classify help desk tickets based on natural language notes. These advances take advantage of the combination of vast computational power, algorithm development and practically limitless data storage.
Consider how machine learning, combined with Siri or other voice-activated applications, can help you plan a night out, make a dinner reservation, buy movie tickets or map a route to a destination. In the smart infrastructure context, using collaboration technology and machine learning, city workers can identify a water leak, potholes in a road, and use their smartphone and machine learning to report the problem, order replacement parts and more. These integrated technologies work together to enable new kinds of efficient outcomes.
To better understand both opportunities and real-life examples, consider Waze, which is a mobile navigation application. It uses human collaboration and real-time updates, combined with mobile computing, to help users easily navigate and avoid traffic. The technology relies on drivers to report traffic congestion and road hazards. The application supplements human-reported data with centralized compute intelligence to detect heavy traffic conditions, road hazards (potholes or disabled vehicles), and more.
This traffic example represents a kind of “implied” collaboration. Waze determines the “best” navigation path using official maps and its own calculated routes, combined with user navigation behavior. In this way, we blend straight compute, social data input, machine learning and human collaboration to determine the optimal path, which changes as traffic patterns shift. The quality of information is enhanced by collaboration technology, crowdsourced data and intelligent systems.
Robust, compact and portable smart devices have opened a new world of possibilities. Regardless of format, these devices provide foundational technology in the digital age. At a basic level, they are a communication technology. But at a deeper level, modern mobile devices offer a rich set of sensor data, mobile compute for on-site work, and connect workers to powerful compute and intelligent systems.
The most common modern devices are smartphones and tablets. While basic wireless telephony has been common for the past three decades, only in the past 10 years have mobile phones become smart. Today, the telephony component is a mere app among a phone's many functions.
Just consider the vast array of sensors and capabilities in today’s smartphones. Most have a combination of proximity and light sensors, cameras, barometers, gyroscopes, near-field communication, microphones and more.
“Robust, compact and portable smart devices have opened a new world of possibilities.”
Every sensor produces data that can be used alone or in combination. Using an accelerometer sensor, combined with the gyroscope, a developer could detect a phone drop – or a human fall. This last example, if combined with a fast network, additional computing power and enough storage to collect longitudinal data, could also yield an emergency service function for at-risk elderly patients or chronic-disease sufferers. In short, mobile devices today have broad capabilities to solve complex societal challenges.
As mobile devices and IoT sensors proliferate, bandwidth's need for greater speed and the ability to ingest more data only increases. As a result, new computing models such as edge computing are augmenting mobile device capabilities. Edge computing models enable data to reside at the edge, meaning it does not need to travel back to a centralized data center or cloud. With the capability to store and analyze data sets at the edge, practitioners can act on data faster and potentially via new formats, such as mixed reality headsets or smartwatches.
Beyond phones, a range of mobile and nonmobile devices provide additional capabilities. One key trend is smartwatches or activity trackers. Advanced sensor technology allows some of the same sensors in phones to be scaled to watches.
These watches count steps, determine barometric pressure, receive GPS signals, assess attitude and even make phone calls independent of a smartphone. The last feature – telephony – effectively creates a small edge device with an independent data connection capable of transmitting all the sensor data directly to the cloud in real time. This makes use cases like remote patient monitoring or human performance monitoring possible. It also brings an opportunity to ensure the health of chronically ill patients and to modify training programs as conditions change. Further, all data can be available to healthcare teams or coaches as quickly as the data can be transmitted to monitoring software.
Advances in computational power, virtually unlimited storage, fast and inexpensive network connections, and powerful mobile devices form the foundation for greater innovation. Still, human collaboration is necessary to capture any benefit. None of the data or connectivity means much without multiple individuals working collectively to overcome a common challenge or achieve a unified aim.
Let’s look at two practical examples, starting with public safety.
Command and control for first responders is increasingly mobile. Body cameras, combined with sensors for gunshot detection, weather, video feeds from city and commercial buildings and social media platforms provide a robust data set that can be analyzed to improve public safety. Data can be instantly relayed to first responders to provide real-time updates on developing situations – from fires to acts of terrorism to routine public events. Cloud and mobile-based compute can provide insights to identify ideal traffic patterns and public safety hazards or to assist in personnel distribution. However, as in healthcare, the key is the ability for humans to use these insights collaboratively and collectively – often in real time.
Heart disease is the number one cause of death in the world. Technological advances, specifically telemetry ingestion through monitoring and large data set analysis, can detect anomalies that might provide early warning of a cardiac event.
A Chicago-based firm has taken technology used in jet engines and applied it to humans. Combining the technological factors we’ve described here–network connectivity, mobile devices, cloud computing, machine learning, and cheap storage–the firm has developed analytics based in machine learning that apply artificial intelligence to physiological data collected from bio-sensors. With this data, the tool can learn relationships and patterns in a person's vital signs. Importantly, this insight is used simultaneously by healthcare practitioners to create a collaborative treatment plan for patients. In fact, collaboration between these parties enables better healthcare and, presumably, better outcomes for patients.
“Technology won't solve specific problems without the human element, though. ”
Combining collaboration with smart infrastructure has also created new environments enabling humans to make better use of the information they receive. In cities like Boston, citizens can report problems or needed repairs by sending a photograph of the issue to a mobile app maintained by the city. Images and conversation spark conversation among these citizen activists to identify improvements using real-time tools and technologies.
Los Angeles has installed smart street light technology that does more than illuminate an area with more cost-efficient lighting. These smart nodes also support mobile phone service, include surveillance cameras, and charge electric vehicles.
But Boston’s mobile app also highlights a key point as cities and organizations strive to build smart infrastructure: No amount of costly smart infrastructure will make a city intelligent unless smart infrastructure technologies are successfully combined with communication and collaboration technologies and processes. “Too much investment in infrastructure and not enough in collaboration, particularly with users and citizens, won’t make cities any smarter,” wrote Harbor Research.
Technology has provided tremendous opportunities to combine information and make experiences smarter. These advances, however, won't solve specific problems without the human element, though. Organizations should strive to choose technologies that can integrate with one another and that can provide easily digestible data that humans can then act on, in the moment. Without these capabilities, collaboration tools and smart sensor technologies can’t reach their full potential to improve and make processes more intelligent.
Selecting technologies that are efficient, intuitive and cost-effective requires humans to think about the environments in which this data will be captured and used and how humans need to interact with these tools to get the most out of them. The buying process should put these tools to the stress test and require them to work together to yield information. When stakeholders work together to identify their key needs, they can devise appropriate technology solutions, not buying something because it’s the next bright shiny thing.
The key is collaboration.
Sean Shell is a collaboration technology expert and consultant.
Connect with him on Twitter at @shawnshell.