If data is the foundation of a smart city, video is the cornerstone. It is already ubiquitous in many areas, and with smart video systems, cities are doing everything from managing traffic to identifying opportunities to improve their businesses.
But video delivers the greatest value when it is integrated and analyzed with other data sources, like Internet of Things (IoT) sensors. Combine traffic cameras with pollution sensors, for example, and a city can form new strategies for managing air quality.
More broadly, data integration can help automate workflows, aid law enforcement, and improve delivery of city services. Layering video on top of other information can help experts, managers, and officials better understand the big picture and the dynamics of how systems affect one another. Add artificial intelligence, and the combined data can be used to identify patterns and trigger alerts for scenarios that otherwise might go unnoticed.
Challenges of Video
Although video offers strong potential, some deployments fall short for three reasons.
For all of video’s potential, it is cumbersome for human beings to process. To fully understand video content, people have to watch it, making the process expensive and time-consuming. It is often used after an event occurs for evidence and reactive, not proactive, responses.
The second challenge is data size. Video files and streams are enormous. Most cities lack specialized IT infrastructure designed for video transfer, storage, and retrieval. Accessing and processing video through existing networks can be a slow process that dominates network resources, interfering with all other operations, and even sometimes leading to data loss.
The third issue is the siloed nature of data. Cities typically contain a plethora of IoT and other data-generating systems, each designed to operate independently. These systems rarely share data formats or interfaces, making it difficult to combine their information.
The result is that for all the available video, almost none of it has been used at all, let alone to its full potential—until now.
New solutions like Hitachi Smart Spaces and Video Intelligence and the Hitachi Visualization Suite (HVS) integrate and intelligently manage much of a city's data. They can mine video for important information, transform it into insights and alerts, integrate it with other types of siloed data, and provide a holistic picture to help improve operations and safety.
Integration with Disparate Data Sources
HVS uses Intel® technology to generate, analyze, and bring together many types of data — video, environmental sensors, social media, geographic information systems (GIS), weather feeds, traffic reports, and more — onto a single pane of glass (Figure 1). It intelligently connects to a city's existing software infrastructure to enable an array of new applications.
An analogy would be driving a car. When on the road, you use your eyes and ears for clues to what other drivers are about to do. Side and rearview mirrors supplement the visual information from the front and rear windshields. Gauges let you know speed, time, and various types of mechanical status. A GPS or smartphone delivers directions and the radio can alert you to traffic conditions. Imagine having to drive a car while logging into and checking these different systems, plus deciding where to accelerate, stop, or turn.
Similarly, the availability of many data types within a single view helps city officials make smarter decisions. Historical trends and analytics can help drive smarter strategies. Video combined with mapping and other types of data can provide situational and geospatial awareness in areas that include public safety, law enforcement, transportation, and the operation of water, sewer, and other utilities.
Data features are no better than the ease with which personnel can use them. HVS employs browser-based software that can be used anywhere, and an intuitive user interface that combines power with a minimum need for training (Figure 2).
With all the power for users, a city can still maintain central control and establish responsible governance with enforced separation of controls. The multi-tenant system provides for user-specific views and permissions.
The results are improved productivity that can drive fiscal and operational improvement. On average, public safety could improve at twice the current rate while the pace of road transportation improvements could be 20 percent higher than now. The increase in service levels has the potential to make our cities safer, more efficient, and better places to live.
For example, intuitive investigative tools and simple search capabilities, timeline views, and Digital Evidence Management (DEM) can help various aspects of city management. Whether law enforcement or risk management, they can more effectively undertake their work.
Rule-based workflow automation makes responses more proactive, and brings additional efficiency, taking actions and sending alerts to appropriate personnel based on events.
Analytics Transform Video Data
Computer vision and machine learning transform video into a wealth of insights and alerts. Rather than waiting for humans to review video, AI automates many aspects of video analysis. The Hitachi solution can perform a wide variety of tasks. This includes counting bikes, cars, trucks and public transport, observing pedestrian flow, detecting left-behind objects, applying face recognition, or identifying a specified type of event and sending an alert to the proper officials.
This video analysis combined with predictive analytics helps to identify potential problems before they occur. Implementation of machine learning and computer vision, based on Intel technology, reduces the burdens of central data management and analysis.
Real-time video can be available when necessary, but with intelligence at the edge, cameras can send just the insights or alerts, relieving data transmission networks, reducing fragility and costs. Specially outfitted cameras, or existing cameras coupled with edge gateways, can apply analytics to transform visual information into metadata that's smaller and more manageable.
A camera might analyze traffic, taking counts of different vehicle types and sending the counts, instead of streaming video back to a central server to be viewed or analyzed. Or it could integrate facial recognition at airports and major ports to send alerts when it recognizes suspects or missing persons.
For example, Las Vegas is using HVS trial applications in a neighborhood designated as an innovation district. The city is analyzing traffic to count bikes, cars, trucks and buses, and flows of people through its transit stations, as well as parking occupancy.
Integrated data analytics dashboards can show operations personnel and urban planners when transit stations are busiest, how drivers use parking spaces throughout the day, and where patterns of foot and bike traffic occur — all factors important to businesses.
One surprising realization was that bike traffic was far heavier on one path than predicted. The data led the city to speak with local businesses. Merchants reported that a significant number of cyclists were picking up and delivering restaurant orders. The insight helped justify installation of more bike lanes for safety, better overall traffic control, and improved commerce.
Austin is another example of a city that has applied HVS to its public safety functions. Law enforcement agencies had siloed data that limited their visibility when incidents occurred. Now the city has access to video monitoring, computer-aided dispatch systems, license-plate and facial recognition, gunshot sensors, radioactive isotope detection, and other technologies that come together on the screen to better inform city officials.
HVS enables the city to expand use of data-driven crime control. The system has pulled the data together for Austin so officials can perform real-time monitoring of criminal behavior and better emergency response.
The police department can deploy officers in areas where crime is more likely to happen, thanks to predictive analytics. Officers dispatched to an incident also have more information to help them determine strategies and tailor their responses to the situation before arrival.
One of the added benefits to Austin has been the ability to integrate third-party cameras of businesses willing to share their data. The capability saves taxpayer money while it expands visibility and builds stronger relationships with the local community.
By making video and many other types and layers of data readily available to personnel and planners, HVS and technologies like it provide important tools to cities. Municipalities can now recapture and expand the usefulness of data they already collect without a major restructuring of existing systems.