CitySight is a new cloud-enabled performance management system — another example of innovative Xerox data analytics, and a way of combining machine learning and human intelligence.
Computers are really good at churning massive amounts of data sets and finding patterns and trends. Data analytics help us do that very efficiently. However, as many data scientists know, it’s not always easy to apply “computer logic” to solving complex problems that come up in our chaotic world.
The computer is only as good as the data it has access to. And it is challenging to create systems that can extract the knowledge needed to make decisions out of data that is coming from so many sources and sensors – including video, social media, documents, voice. And it’s equally difficult to allow room in the system to utilize one of our greatest real-time sensors available: humans.
According to Mark Stefik, PARC research fellow and project lead, humans excel at classifying and seeing patterns in very noisy and imprecise environments. (Like walking city streets to monitor parking enforcement.)
While there are machine learning and artificial intelligence (AI) tools under development to help humanize computer decision making, Stefik and his team are focused on how humans can serve as living breathing “data” sensors. Essentially, CitySight can take into account the real-time data that officers see, and use it to make the best decisions in the moment.
PARC scientists are exploring the creation of what Stefik refers to as “extreme teams.” The combination of human intelligence (real-time, on-the-street officer input) and computational expertise (the computer and data analytics) results in better, more efficient decision making. Without real-time data analysis, today’s teams have to resort to manually reviewing data to spot trends and visual patterns, which takes time. More often than not, they simply make decisions without having the benefit of the data.
The innovation behind CitySight was born “on the street.” Xerox data scientists and skilled ethnographers spent hundreds of hours watching the daily activities of parking enforcement officers.
CitySight analyzes the daily activities of parking enforcement officers (PEOs), such as time worked, patrols, citizen interaction, and citations. Supervisors use the insights gleaned from the data to manage time and resources to improve the flow of traffic, ensure the availability of parking spaces and keep citizens safe.
The innovation behind CitySight goes beyond data analytics. It’s the convergence of cloud computing, smart mobile devices, big data analytics and social media.
AI is built into the analytics as human intelligence enters system via structured messages from the officers, which are combined with other data (GPS and other sensor data). This facilitates computer understanding of the human input in real time.
The Xerox innovation team has their sights on more than just CitySight. They see promise for this approach and are developing a menu of system elements that can be combined and deployed in a variety of forms. This can help municipalities, hospitals, governments, and big and small companies, make better decisions by combining the right mix of human and computer decision making.
Stay tuned for where this technology will enable its next extreme business process make over. It could be in your industry!