• Smaller Small Medium Big Bigger
  • Default Helvetica Segoe Georgia Times

By Kamal Ballout, head of TEPS MEA & Global Energy Segment, Nokia

In comparison to the Industrial Revolution, the Information Revolution has been disappointing from a productivity perspective. But all of that is set to change with automation trends such as smart cities and Industry 4.0. The digitally driven automation of our infrastructure will have very productive results in many areas of life, including manufacturing, energy, transportation and many more. IoT married with advanced technologies such as artificial intelligence (AI), cognitive/machine learning and machine-to-machine communications will finally enable us to harvest the fruits of our digital data for real-world results. Many things are coming together to make this possible, but at the center of it all is the network, which is the real driver of this revolution.

For example, 75 percent of the workforce in the US works in asset-intensive organizations such as manufacturing, power utilities, mining and transportation. But these industries have seen very little growth in comparison to the digital sector, which has seen four times the level of growth since 2000. These industries have employed digital point solutions, such as robots in manufacturing and mining, smart grid technologies in the power sector and advanced signalling systems for railways. While effective, these isolated systems could be much more productive if integrated across the organization using high performance networks, cloud technologies and cognitive analytics.

IoT, the key for optimal productivity
For all of these sectors, the availability of IoT sensors connected to the network is a key part of the shift. Take for instance, asset monitoring, which is critical to controlling costs and ensuring reliability and up-time of systems. Current maintenance techniques, which use statistical measures such as mean time between failures (MTBF), can be much more accurate using networked sensors and machine-learning. Instead of relying on historical data and average statistical models, machine learning uses the real-time data from embedded sensors to spot anomalies. This allows for predictive maintenance (PdM), which is much more accurate, thus saving time and resources that are either being spent on over-maintenance or reactive maintenance. Employed across the cloud, these PdM systems can learn from thousands of similar machines being used anywhere in the world, making them even better at predicting failures.

Since many of these organizations have complex and distributed functions and processes, communication between functions is critical to improving overall performance. In the era of renewable energies, as an example, energy grids are now much more complex and are required to do more than simply distribute electricity to users. With solar panels on their roofs, households are also now electricity producers. In telecom terms, the energy grid is now “full duplex” and, as a result, far less predictable. Integrating the diversity of distributed energy resources and dynamically maintaining consistent, predictable power for everyone now requires more intelligence and better communications.

In manufacturing, robots have been heavily employed on assembly lines for years. Automated guided vehicles (AGVs) follow pre-set magnetic tape to deliver goods between different parts of the line. But there is little or no communication between them. Re-tooling is very manual and can take months and years instead of days and weeks. To become more flexible and responsive to rapid shifts in consumer demand, leading-edge factories are creating greater autonomy and mobility for robots, arming workers with augmented reality headsets and smart tools, and employing self-driving AGVs that flexibly collaborate with others to meet shifting needs. Sensors are everywhere, monitoring manufacturing processes for quality and the local environment for hazards.

Networks are the nerve centers
Getting all of this to work as a coordinated whole depends on mission-critical mobile communications that replace Wi-Fi and Ethernet with LTE (and 5G in the future). The results are dramatic: more flexible and faster re-tooling, better safety and increased productivity as all the parts of the factory now communicate in real-time. AI and machine-learning analyze this streaming data and use it to suggest better ways to organize flows, spot anomalies that might lead to hazards or quality issues, and provide augmented intelligence to on-floor personnel to improve performance.

Fields as diverse as mining, retail and financial services are also experiencing this shift, as are intelligent transport systems (ITS), health care and public safety, which are all part of the much larger and growing field of smart cities.

The need for integrated platform for automation and smart cities
Cities are a good example of the automation trend. Like industrial environments, they are asset-intensive and face many of the same challenges. They need to be able to predict mechanical problems within municipal infrastructure of all kinds, from waste-water processing plants to traffic management systems to street lighting installations. Managing these vital processes, requires enormous quantities of data from a vast number and variety of sensors, actuators, video cameras and other connected devices. Interconnected with a high capacity, secure and reliable network and utilizing localized computing resources or edge computing to provide rapid processing capabilities, these sensor and camera networks not only provide meaningful information to city managers to respond quickly and intelligently when issues arise, but they can also help to automate much of the routine maintenance and many daily operational tasks, creating time for municipal staff to focus on other priorities.

Smart cities make for an interesting example of how this smart network fabric, connecting sensors, cloud processing and cognitive analytics, operates best as an integrated platform. For instance, cognitive analytics capabilities that are used for PdM can be shared across many diverse functions. Today, the transit system probably uses a different management system for asset maintenance than the waste-water facility or parks and recreations. But the cognitive analytics capabilities can be shared. Similarly, sensor and device management are probably siloed today, but can be shared by multiple functions.

This integration is about more than sharing the cost of these complex systems. The data from all of these integrated functions is much more valuable when combined than as separate data lakes. The real power of AI and machine-learning is the ability to process vast amounts of data and make correlations in real-time that lead to valuable and actionable insights about how to improve city services as a whole.
An integrated platform can also be opened to partners and citizens to encourage innovation. Smart cities, like many public institutions, are both budget-constrained and tend not to be good at innovating. Many of them are pursuing open data initiatives that expose selectively the data they collect and analyze making it available to others to innovate better services for everyone.

This platform approach to IoT, data and analytics only works when it is supported by a high performance network and distributed cloud infrastructure. The adoption of video is driving the need for much higher bandwidth, and the mission-critical nature of many of these functions demands telco-grade reliability. Many automated functions require low latencies that only edge computing can provide. Many cities and highly distributed enterprises will turn to partners to provide them with these capabilities.

We are on the cusp of a revolution that will utterly transform life as we know it. The results will be profound across virtually every sector of society. The joining of information technologies, the cloud, IoT and various forms of automation is all being made possible by the network. That makes for an enormous opportunity for today’s network operators.

Pin It