The Road Toward Behavior-Driven Automation

Andre Fuetsch
2021 IEEE wireless communications  
6 If 2020 taught the world something it is that adaptability is essential, and flexibility is key to social and economic stability. In times when health safety is paramount, with small IoT devices producing unimaginable amounts of data and automation driving productivity, the world is on the evolutionary path toward a more integrated knowledge-driven society. The next chapter of technology evolution is bringing together existing technologies such as the Internet of Things (IoT), data analytics,
more » ... artificial intelligence (AI) and 5G, making it possible to enact human behavior changes and enabling context-based decision-making to augment worker productivity as well as safer and healthier communities. The COVID-19 pandemic accelerated a global shift toward a mobile, location-agnostic society and the rise of a nomadic generation. Social distancing does not mean social isolation, and the need to understand user behavior is more critical than ever. From co-working to remote work or just to collaborate and interact with business partners and friends, it is becoming quite clear that the trend is here to stay and it opens up a new way of looking at perimeter security as well as the need for ubiquitous seamless connectivity. Behavior-driven network selection is essential, and users should be presented with access network choices based on their "persona" allowing them to securely switch between work, study, or entertainment. Seamless broadband fiber and wireless network integration in the home and office will allow them to share a printer, pause streaming movies or control smart devices. This seamless network access comes with challenges, especially in the shadow of several high-profile cyber security attacks on national infrastructure. Hard wired, rigid security policies are not effective in hybrid working or learning environments where devices are moving in and out of enterprise networks. This demands a new approach where user-based context-awareness allows applying zero-trust policies across multiple network domains. Zero-trust is a paradigm shift in network security and assumes that the network has already been breached, therefore requiring authentication and access control checks as users access internal systems. This type of authorization needs to be frictionless and effortless on the user's part so companies are looking at multifactor, password-less authentication mechanisms coupled with context-awareness checks such as location, device or even the access network identity where the user is connecting from. And it is not only for network security; all the context and behavior awareness are driven by data. Data is also becoming more critical than ever from defining health policies to evaluating human impact on the environment. While data is an essential ingredient in process improvement and progress in general, the sheer amount that is generated via technology makes it unusable in a raw format. Through data analytics, we turn data into information, adding structured ways to evaluate and use it. Information-in-context becomes knowledge, and when combined with behavioral psychology, generates wisdom. As a result, a new trend is emerging and is expected to grow in the years to come. Enter the Internet of Behaviors, a concept first coined by Gote Nyman, a retired professor of psychology at the University of Helsinki in 2012 and elevated to stardom by Gartner as one of the emerging technologies of 2021. Internet of Behavior makes use of the data generated by the vast IoT networks of sensors and smart devices to influence behavior. It has the potential to change the way we use technology, how we share data, and more importantly how we look at privacy. At the very high level, IoT is a network of networks of devices and sensors under various administrative domains. Internet of Behavior brings user centricity and adds "human nodes" into the mix to combine and correlate data from devices with the power of social networking. The ability to gather, aggregate, and utilize data to influence or analyze behavior is the new way of interacting with customers and employees. By attaching behavior to a user's profile, companies can provide a better user experience in retail, transportation and other industries, while allowing them to proactively identify and resolve issues. It should enable marketers to better understand customers and offer relevant products and services as well as empower users to make health, safety, or product decisions by placing data-driven options in their hands. The concept of altering behavior using data is not new. Probably one of the simplest examples is the use of speed displays on specific accident-prone spots on roads. Chosen locations are already the result of data analytics collected using smart city sensors aggregated with weather information and vehicle demographics. Alerting the driver of their speed and making that data highly visible on the road proves to be effective at enforcing the speed limit by leveraging behavioral psychology. Computer-vision, coupled with data analytics, can drive the enforcement of safety regulations such as the use of masks, and it can be used in conjunction with thermal sensors to track peoples' temperature in public places such as malls or stadiums as well as in the workplace. There's no better example of an expected behavior-altering application than police body camera solutions. Another computer-vision use case is applying facial recognition for customer demographics analysis used in conjunction with location and behavior aware preference selection for consumers. These are also good examples for realtime, low-latency and privacy sensitive applications that call for deployments as close as possible to the data-sources, making these prime candidates for Edge Computing. Edge Computing brings compute, storage, and memory resources closer to the Edge of the network, allowing the offload of latency sensitive workloads from the cloud data centers. Due to privacy sensitivity, typical solutions would have to process raw data, anonymize, and aggregate without storing or transferring any individually identifiable data. Health applications are another great example of user-driven behavior-changing solutions. Unlike previous use cases, the user is the aggregating hub of raw data from sensors such as scales, blood pressure meters, fall detectors in addition to location data, environmental data available from municipalities on air or water quality to allow the users to adapt and adjust their behavior toward improving their health. A special category of use cases are collaboration robots or cobots. These are a new generation of robots designed to help in the automation of processes. Unlike typical robots programmed to perform a very specific task such as painting a car or attaching a windshield, their specific goal is interacting with human counterparts, understanding their behavior, and helping them by increasing safety and productivity. One example, for instance, is a cobot that is designed to enter toxic areas, lift heavy objects, or bring parts to a team of workers. We can think of them as non-human apprentices. While manufacturing is a significant sector benefiting from their applications, other industries (retail, logistics, etc.) are adopting collaboration robots and they are expected to take advantage of the deploy-
doi:10.1109/mwc.2021.9535458 fatcat:ryopljpnwfafzlxhoxpqki5vvm