iHotSpot is a new-generation predictive analytics engine capable of providing high-accuracy predictions with much less data and using shorter computing time. For example, this system involves the application of a pioneering AI algorithm "Network-based Graph Deep Learning (NGDL)" in the processing of a wide range of location-tagged and time-stamped data, including traffic accidents, traffic flows, crimes and fire incidents.
Here we use open crime data to demonstrate iHotSpot. This system allows the user to access Chicago's (historical) crime hotspots predicted one day ahead of the day chosen by the user, for a crime type (burglary, theft and assault) at street level based on a risk level chosen by the user. The risk level is measured by an 1-15 band reflecting in which top percentage the area is most prone to crimes in the city. The predicted hotspots can be contrasted with the crimes that happened on the day.
iLogistic provides tools for managing, planning, and optimising the supply of public- and private-sector services in urban areas, based on the spatio-temporal patterns of service demands. Our product aims to improve the efficiency of services by utilising the most appropriate spatial scale (e.g. street segments, urban cells, or census units).
Strategic design of service district (e.g. police patrol areas, delivery zones)
Route planning of service units (e.g. police patrol routes, delivery routes)
Location selection of service points (e.g. shops, shared bike stations)
Management of real-time demands (e.g. emergency response, fire response)
Rebalancing supplies (e.g. daily rebalance of shared bikes)
iLogistic has been successfully employed in a range of urban services. It has been supporting the strategic and tactic planning of police in Greater London and Kent.
iLogistic supported the location selection of docking stations and the daily rebalance of station-level bikes in Shanghai.
iMode is a holistic space-temporal product for providing deep insights of transport operation, including multi-mode traffic flow (car, bus, train, tube/metro, cycle, walk, stationary) and active travel profiling. It is developed based on our advanced travel mode detection algorithms and comprehensive mobile GPS datasets. iMode can measure the carbon footprints for local communities, achieving Net Zero Target in both urban and rural areas.
iMode is useful for local authorities to evaluate the impacts of Low Traffic Neighbours (LTNs) on traffic flow and travel behaviour changes.
Our flexible bespoke framework and service could also provide products in various spatial (road segments, OA, MSOA, LSOA, Local Authority) and temporal (hourly, daily, weekly, monthly and yearly) units.
iPlace is the finest space-temporal insight for places and areas, providing deep insights of human activity (visiting patterns, footfalls) and mobility (flow, migration) in big and small areas, cities based upon advanced space-time analytics, which can support business and government for place-related decision making.
Place footfall patterns (support bespoke POIs, e.g., Tesco, Perit, Costa)
Retail impact index ranking (support MSOA, Local authority/borough area analytics)
Retail recovering index (support MSOA, Local authority/borough area analytics)
Urban area footfall patterns (support LSOA, MSOA, Street segment, LA)
Commuting/migration Flow between urban areas (inter-city or intra-city)
High street vatality