“For intelligent transportation, there needs to be a near-human level of detection. The human reflex arc is two milliseconds. If you automate this detection into a car, it needs to be under five milliseconds.”
This is how Professor Vincent Chan, the Joan and Irwin Jacobs Professor of Electrical Engineering, Computer Science and Aeronautics at MIT, led off his team’s discussion of how improved sensor technologies are key for the future of smart city transportation systems.
Professor Chan is working alongside team members Soung Liew of CUHK, Victor OK Li of HKU, and Danny Tsang of HKUST to create a baseline IoT network architecture for intelligent transportation systems centered on:
The team’s plan is to leverage this architecture to develop cognitive networking for smart cities in order to create more timely communications for signal control, freeway management, even more-efficient public transit, and enhanced public safety.
The reasoning is simple: Safety and security. Not only for those driving on the crowded city roads, but also for those in need of speedy assistance – an all-too-common problem in today’s city infrastructures.
A truly smart city infrastructure consists of multiple components that require the timely exchange of information within two IoT-enabled services: critical messaging (CServ)1 and baseline communications.
To this, Professor Chan said, “Current systems, such as WiFi, [don’t] have the [necessary] security and time deadline guarantees creating issues with emergency response. We have concerns and need to make sure that critical communications across systems can happen at an extremely high level of reliability and timeliness.”
The basis of their approach for smart traffic management comes from a somewhat unlikely source – the networks of the Defense sector. The defense R&D agencies recently studied the use of sensors to provide reliable, time-sensitive communications over unreliable networks, accommodating bursts of user traffic through dynamic mobility, and then garnering accurate analytics to improve delivery.
Using this general approach as a baseline for its IoT network research proposal, the team believes it can deliver an architecture framework, alongside the algorithms necessary to create a heterogeneous communication system by the end of year one.
In turn, to develop cognitive networking for IoT and Smart Cities, the research will focus on multiple challenges.
• Inferring the state and reliability of the network based on actual traffic, combined with performance statistics based on active probing.
• Inferring the confidence on the network's state based on sparse data sources while also sometimes using stale data.
• Making network design decisions based on load balancing, reconfiguration and restoration requirements.
• Prediction of user intent, leading to appropriate action and network response in a timely fashion.
As Schneider Electric continues to make EcoStruxure a vital tool for IoT implementation across its market segments, this unified research initiative between academia and industry is crucial in the development of cutting-edge research of sensors, signal processing and analytics for building and transportation connectivity.
1CServ, Critical Service is a delivery time deadline guaranteed network service.