NEC develops IoT technology for responding to environmental changes and disasters
Via NEC News room
Jul 17, 2015
Tokyo, July 17, 2015 - NEC Corporation (NEC; TSE: 6701) today announced the development of an "autonomous and adaptive control technology" based on the Internet of Things (IoT) for dynamically allocating resources in response to unpredictable environmental changes and disasters.
IoT enabled systems for managing social infrastructure encounter a variety of unpredictable conditions, such as intense weather, disasters and accidents. However, it is difficult to predefine effective rules for ensuring that systems function at their best possible level when exposed to such uncertain environments.
In order to enable social infrastructure to properly respond to unpredictable events, NEC developed an autonomous and adaptive control technology inspired by the adaptation processes of living organisms to environmental changes. One of the key features of the technology is its ability to perform dynamic optimization control using only prescribed indices for a target state such as efficiency, comfort and safety. Moreover, this is accomplished without requiring predefined rigorous control rules for adaptation.
NEC tested the technology by conducting real-time simulation of the best allocation of taxis in response to customer demand, and was able to improve customer acquisition by more than 10%. In energy management tests conducted in offices and houses, use of the technology led to approximately 10% energy savings, without sacrificing the comfort level of tenants.
"In addition to helping adjust for supply and demand in taxi ridership management systems and the improvement of energy utilization efficiency in energy management systems, the technology can also assist with a wide range of applications, including the agile response to sudden cargo pick-up demands in logistics, the efficient allocation of security guards in public surveillance for safety and the mitigation of congestion in overcrowded areas. Going forward, NEC will also pursue the development of actuation (control and guidance) technologies using big data," said Motoo Nishihara, General Manager, Cloud System Research Laboratories, NEC Corporation.
Key features of the new technology include the following:
An asymptotic approach to reaching a global optimum state without predefining control rules, even under highly uncertain environments NEC developed a unique method (algorithm) that divides an entire system into subsystems and enables each of the subsystems to dynamically and autonomously decide its own operations. The method assigns individual indices for each subsystem based on the control targets for the overall system and decides rules and operations by comparing individual indices among subsystems. This allows optimization and improvement of efficiency of the entire system, without the need for predefining control rules.
Real-time adaptation of large-scale systems to environmental changes Distributed calculation by each subsystem using only localized information enables streamlining of computing operations. Unlike methods that carry out intensive calculations using entire system information, the technology makes it possible to repeatedly perform optimum operations at short time intervals as well as respond to unexpected environmental changes in real time, even for large-scale systems made up of a large number of devices.
High fault tolerance in response to IoT data deficiencies and system failures Despite the occurrence of data failures or deficiencies in some subsystems, the entire system can continue to function at its best operational state possible using only the normally functioning subsystems. It therefore realizes a high level of fault tolerance to enable continued operations and withstand the occurrence of unforeseen failures in some parts of the system.
Stay up to date with the latest industry developments: sign up to receive TelecomTV's top news and videos plus exclusive subscriber-only content direct to your inbox – including our daily news briefing and weekly wrap.