I imagine a future of pyramid buildings, so everybody gets a terrace with a vast sweep of sky, and so we have more green with the building than would be on the ground without the building. There’s a network of…
I imagine a future of pyramid buildings, so everybody gets a terrace with a vast sweep of sky, and so we have more green with the building than would be on the ground without the building. There’s a network of sensors zipping data on soil moisture and air quality back to the brain of the city, joined by data from all manner of other things – like electric (or pedal powered) shared transit that has no need for a left nor a right hand lane – the vehicles swarm like swallows.
We can be smart with each of these data sets. But what if we want to explore the relationships between the data sets? Maybe we’ll find something interesting in a cross reference of the movement of people, with air quality, and geographic data on health – just one example.
Across the city, there’re multiple IoT platforms serving different applications with different technologies and standards. This fragmentation makes it difficult for applications to talk to each other. To derive full value from IoT investments, open standards are needed. Lighting, parking, security, the grid, all need to be talking the same language. Municipalities and businesses all need to be on board. With an interoperable IP ecosystem, there’s a broader range of vendors, meaning pricing is favorable, projects more feasible, and continuity of supply is better assured. Network equipment can also be reused across many applications, and different applications can share data more easily, and interact to enable smarter functionality.
New projects and benefits for citizens will increasingly come from the data we’re getting from existing solutions. Data from ‘x device’ can inform data from ‘y device’, making us wise to ‘z’. But many of the possibilities in these intersecting data sets we’ll know nothing about unless ‘x’ is talking the same language as ‘y’. With devices all talking the same language, the possibilities are now limited only by our imaginations, or by the imagination of AI, when we feed it with:
Environment data: With data on emissions and airflow, AI predicts pollution levels, and recommends measures to improve air quality. This is an answer to particulate matter (microscopic solid or liquid particles hanging in the air) which is damaging to the lungs, and a problem in many cities. AI can also monitor and help ensure water quality, and detect energy wastage, recommending measures to save energy.
Transport data: AI optimizes public transport routes, which reduces costs, and gives users more accurate arrival times and a better experience. Different networks – buses, rail – are integrated into one app. Deep learning algorithms predict traffic build up, and recommend alternative routes to private drivers, to prevent congestion, which also cuts down on emissions. Lights stay green longer when they need to, to ease traffic. AI enabled cameras and traffic violation systems help reduce accidents.
Parking data: IoT and AI informs drivers where they’ll find a parking spot. A nicer experience, and no belting out of fumes driving around the block looking for a spot. Ticketing processes are more efficient, too.
Waste management data: Trash never used to be anything exciting; ‘tis these days. Sensors monitor levels, making collection more efficient, and strolling down the street we no longer have the not-so-sweet aroma of overflowing bins. AI also makes more efficient the sorting of recyclables.
Water management data: Sensors monitor water levels in tanks, condition of pipes and pumps, water pressure, and groundwater levels, to optimize water management, improve water quality, and save water with repair personnel getting to leaky pipes sooner.
Security data: Static cameras and drones detect criminal behavior, and identify and track criminals.
Many projects within these categories are well underway:
MachineCanSee’s neural network understands vehicles and roads in three dimensions, building a vector simulation, tracking and predicting vehicle movement. MCS build customized projects with their technology.
CityZenith is an open source urban twin platform. Users can build server-side and client-side apps to automate tasks, process data, and inform decision making, or hire from a network of service providers.
Clarity Movement sensors collect data on air composition, to help cities understand and tackle air pollution. Modules measure ozone in the air, and enable a better understanding of the composition of particulates; a wind module helps users determine the source of air pollution.
Devic Earth emits pulsed radio waves to increase negative charges of particle pollution, causing the particulate matter to fall to the ground.
Clean Water AI’s deep learning neural network detects and classifies harmful bacteria and particles in water. Cities can install the device across different water sources, to monitor water quality and contamination. Users can see their drinking water under a microscope, as footage like that from a security camera.
Upciti uses AI to analyze images and optimize the movement of goods and people. Examples: analyzing behaviors to improve traffic flow, and detecting free parking bays. Their technology also adapts lighting for users’ safety and comfort, and they create sound maps to identify the sources of noise to help ensure residential areas are tranquil.
Citymapper provides a number of solutions around mobility and transportation, including tools for optimizing transport network design, and operations technology to run intelligent transport systems.
OptiBus’s machine learning and algorithms help make public transit more efficient, sustainable, and equitable, with route planning, vehicle scheduling, and rostering.
BlindSquare seeks to enhance the independence of blind, deafblind, and partially sighted people, painting a picture of the world with sound. Algorithms determine information that may be most useful to the user; shake the device to hear your current destination; mark your position to find your way back later.
Sentry AI extracts data from visual images to detect and monitor people and vehicles; the technology recognizes unusual behavior, aiding the safety of tenants and communities, notifying of intrusions and events such as someone falling.
Nordsense measures the level of trash in a can with 16-point 3D optical laser. An app enables dynamic event-driven collection.
ARcubed filters recyclables from not-recyclable stuff. AI-powered computer vision recognizes waste at the point of collection, and the product gives users feedback to help them make better recycling decisions.
NoTraffic is about solving traffic congestion. Their traffic management platform analyses and predicts traffic flow. The platform can autonomously optimize traffic signaling.
Hayden AI’s onboard sensors and computer vision help municipalities create smarter fleets to protect bus and bike lanes, keep school zones safe, optimize traffic flows, detect highway violations, and improve operational efficiency of garbage trucks.
IntelliVision AI video analytics has a number of security applications, including movement detection, intrusion detection, crowd detection, object left/removed, vehicle or people counting, face and license plate recognition.
Sensus make gas, electricity, and water distribution systems smarter. Smart meters and analytics help utilities and cities monitor such things as usage, pressure, temperature, level, flow, and status, in order to optimize the systems.
Telensa smarten streetlighting. Different lighting policies can be set for different zones, dark skies where lighting isn’t needed, responsive lighting where it’s needed for safety. This plus other efficiencies, in particular the fact of newer LED lighting, gives cities cost savings in the region of 60%, also reducing carbon emissions.
Many people used to believe that AI would take the blue collar jobs first, and never really evolve to be creative. But the reality is looking very different. Exploring interrelationships between datasets from applications such as those above, to figure out ways citizens can benefit, is a creative job, and AI is good at this type of thing. I don’t expect AI to take the creative jobs, but rather be a creative partner to humans. We need the help of AI to work our way through all that data, find patterns, and find new solutions. For that, we need all the nodes to be talking the same language – Internet Protocol.
This is the base on which solutions can continue to evolve in the years and decades to come. An open standards, collaborative ecosystem supported by a certification program for interoperability is vital for effective data governance and seamless interworking. This is how the smart city can flourish. It’s one thing we’re laying the groundwork for at plgd, with our open source, open standards IoT platform built for interoperability.