Machine Learning (Command Centre)

The creation of a Smart City is the melting pot for a multitude of complex and inter-dependent problems. This is where city authorities are expected to step in and proactively monitor/manage the city’s infrastructure, via various information and communication technology channels to ensure a better quality of life for its citizens. Multiple utility monitoring of the infrastructure of a city i.e. the monitoring of water management, automatic waste collection, traffic management, security and surveillance etc. till date, has no readily available singular hardware/software platform; the need for which, is imperative. Thus, the planning of a city’s infrastructure monitoring/maintenance can be done, by the integration of all the utilities on a singular platform called the ‘City Command and Control Centre’, which is a collection of various technologies and engineering functions. Inputs from diverse systems like, electrical, mechanical and hydraulics can be collected by it, by using industry protocols like Modbus and Backnet, along with ICT inputs from Intelligent Building Management system (IBMS) and the varied Information Technology involved in miscellaneous utilities. With mission mode programs like the Smart City Mission in tow, State governments are looking at the deployment of integrated hardware devices coupled to its IT infrastructure, however, the differentiator in this race which sets it apart from conventional system integration is the customized leveraging of machine learning as opposed to just the upgradation of physical hardware. Most OEMs provide inbuilt analytics in their equipment/hardware, which serves only the original design purpose,creating a divide in due technological evolution. The need thus arises, for an agile Command Control Centre, which is flexible enough to allow Smart City/Municipality teams to develop new machine algorithms of their choice via open source tools/techniques like Python, R etc.The creation of a flexible Analytics Platform, hence, shall have the following requirements:
  • A standardized data layer needs to be provided, that gathers and normalizes data from sensors and systems (including external systems/applications).
  • Allowance of Machine Learning algorithms to be developed, by the usage of this normalized data.
  • The development of the machine learning algorithms should not be restricted to only the integrated analytics platform. It should be flexible enough to allow Smart City/municipality to develop, integrate and deploy data science models, using any machine learning tool/technique.
  • Allowing both – the results of the machine learning algorithm to be displayed as a part of the integrated CCC solution and providing flexibility for the results to be displayed on a 3rd party Visualization tool; by allowing the analysis data to be exported into formats like – XML, JSON, Excel, PDF, CSV etc.
For the proper functioning of smart cities, a future-ready, agile, integrated platform that can seamlessly facilitate the flow of information through a flexible, centralized command and control center is crucial; as the citizens of interconnected cities will greatly benefit from the enhanced service delivery/optimization of all utilities. Hence, this paradigm shift, in urban development to be comprehensive, interconnected and agile is essential for the ushering of an urban renaissance, that smart cities promise to pave the way for.

Author: Rajiv Diwan, Practice Head – Advanced Analytics

Rajiv Diwan heads the Advanced Analytics Practice at ITL and is responsible for both – customer acquisition and defining solution offerings of the Practice. He has setup the CoE on Machine Learning from scratch at ITL and has been instrumental in penetrating into new verticals for ITL; including BFSI. Rajiv is engineering graduate from BIT, Bangalore with specialization in Computer Science; having over 18 years of experience in Analytics, Data warehousing, BI and Large Program Management.

Post a Comment