pls 2017: smart street lighting: sensors vs big data
TRANSCRIPT
SMART STREET LIGHTING: SENSORS VS BIG DATA
Keith Henry, AMILPTechnical Sales ManagerTelensa 15 June 2017
Connected streetlights live in a rapidly changing data landscape
• Valuable asset: connected streetlights map where people live and work
• They are ideal hosts for sensors• Central Management Systems (CMS) connect
streetlights to the cloud• But the wider data and application environment
is changing rapidly
Wirelessnode
CMS
• These apps reflect where domain expertise resides• Typically “sense + actuate” applications• Examples includes asset management, traffic planning and
citizen engagement• There is great synergy between applications
Streetlight CMS lives among many expert domain applications
CMS
1 2 3
Big data: new platforms emerging• Connect everything• Analyse the big data• Generate “actionable
insights”• Today restricted to leaders’
dashboards• Potentially huge data comms
costs• Guaranteed ongoing data
integration costs
Big dataplatform
Sensor edge processing: distilled data
• Sensor and processor on the light pole
• Big data distilled into small and useful data
• Small data comms costs
Big dataplatform
Camera-based traffic analysis via CMS
Pole-mounted camera captures video of passing traffic – lots of data. Cameras are reliable commodity hardware today.
Big dataplatform
1
Edge analytics distil useful data
Analytics software on edge processor distils useful info –traffic count, traffic mix etc. Only this small data is transmitted to the CMS. No user-id information is retained, no video is stored.
Big dataplatform
2
CMS traffic-adaptive dimming
The CMS algorithms use the traffic count data to adaptively dim streetlights when traffic is light.
Big dataplatform
3
Rich data shared with other applications
CMS passes traffic mix data with route planning and congestion data with citizen info apps. Open data interfaces.
Big dataplatform
4 4
Big data platform augmentsdata with context
Big data platform takes all of the data, combines it with egair quality and roadworks information, and displays an overall picture on the leadership dashboard.
Big dataplatform
5
©Telensa
Three application layersare emerging in smart cities
1. Sensor edge processingComputing power on the light pole with software for locally processing raw data (eg video) into distilled data
2. Domain expert appsExpert applications that automate departmental activities, such as lighting CMS, asset management, waste, etc
3. Big data platformsCentral systems that combine multiple data feeds to generate and visualisenew insights
What are the Implications for Lighting Professionals?
> Examine potential data costs: transmission costs for sensors and cameras now far outweigh the cost of the hardware
> Consider sensor edge processing to reduce data costs
> Question the justification for a big data platform if the same level of integration and insight can be gained by the integration of expert applications
> Demand open or standard API links between systems, such as Hypercat
> Be vigilant: vendors and integrators of big data platforms often underestimate the challenge and cost of data integration
©Telensa
Lighting will change the economics of smart city applications
> Perfect location: where people are
> Ideal place for sensors
> Powered host for edge processing
> Dramatically reduces the cost of data transmission and integration
©Telensa