In India the median police force deployed per 100,000 is 192, as against the world median of close to 300, as per UN Data. The best way to address this low median is to leverage technology to predict and deploy limited resources more intelligently and optimally achieve higher prevention of crime and thus aim to achieve the tag of an efficient police force.

The turmoil created by Covid 19 the Law Enforcement agencies have been thrust into an unparalleled situation, playing a critical role in halting the spread of the virus, preserving public safety & social order along with tackling the rapidly changing face of crime, new technologies such as AI will be a powerful resource.

AI has the ability to alter the very nature of policing by enhancing efficiency & effectiveness to, for instance, identify persons of interest in crowded places; forecast & predict violence; automatically sort , tag & classify large police operational data such as evidences or harmful materials; monitor drivers of radicalization, are just a few use cases for AI in Law Enforcement Resource Mobilization strategy.

CloudPrisma has identified five main areas for the use of AI for resource optimization:

• Spotting and Mapping of Hot zones – With the collection of historic crime data from various local law enforcement departments, ca be combined with additional datasets, including Socio economic factors, weather predictions, police patrol history and criminological knowledge, to predict crime hot spots in a jurisdiction.
• On Demand deployment of resources – Police resources (personnel, vehicles, equipment etc.) can be allocated based on actual demand of the AI identified Hot Zones/Spots.
• Patrol route scheduling – the use of identified hot spots to optimize patrol routes and schedules.
• Dispatch of resources for calls – dispatching the nearest available resources to respond to service calls based on their predicted response time.
• Response route plotting – identifying the optimal route by factoring in distance and time and then deploying the resource(s) based on availability and optimal response time.

‍

The above AI algorithms can also apply for Law Enforcement Agencies dealing with Traffic, Fire, and Rapid Action Force or Disaster Management Task forces etc.

CloudPrisma brings accelerated use of Artificial Intelligence to develop advance algorithms for different scenarios based on various data sets like –

1. City or Region or Zone based Crime/Traffic/ Record history archive
2. City or Region or Zone based Crime Record recent period (daily/weekly/monthly)
3. Mass Diseases like Covid 19 area wise data like containment zones etc.
4. Municipal data on sanitation, drainage & other utilities
5. Metrological Data with history and correlation with events
6. Deployment and availability of smart sensors, the Internet of Things, next generation telecommunication network (5G, Wi-Fi 6) current or in the future.
7. External Data from local/district/state/national on events (archive and recent period)
8. Area/Location wise concentration of residential, commercial, industrial establishments
9. Any other public data that may have definitive or probable relevance to likely crime or law and order.

‍

‍

Law Enforcement Current Setup

The Law Enforcement existing data across agencies, Other Government agencies, external data, as per requirement. CloudPrisma can assist in identifying sources of relevant data.

CloudPrisma Proposed Solution

• Data Ingestion - data here is prioritized and categorized which makes data flow smoothly in further layers in a common format. Transportation of data where components are decoupled to start analytic capability.
• Data Query Layer - Active analytics processing with primary focus to gather data value so that they are more useful to the next layer. Some of the Tools used are Distributed SQL Query for Big Data (eg. Presto, MY SQL, Appache Spark, MS SQL DB2 etc)
• Analytics Engine – Algorithms & Mathematical Modeling using Machine Learning and apply analytic tools for Statistics, Semantic, Predictive, Text & Video Analytics. (eg. Python, Java, R, etc.)
• Data Visualization Layer - Presenting the data to make it “Valuable” and “Insightful”. Tools using Tableau, Cognos, Power BI etc.
• Intelligent & Secured Hybrid Compute and Storage solution – Providing a mix of On Premises, in the cloud, PaaS and Container based or Serverless computing along with storage types like Object Storage etc.

‍

CloudPrisma strongly feels that the use of AI for resource optimization does require law enforcement to be prepared to answer a number of essential questions during the design process, which will determine the efficacy of the tool. In essence, when it comes to understanding and predicting optimal decisions, such a system needs to know what ‘optimal’ is and how to calculate it. Furthermore, it needs to know when, where and how incidents occur; how resources can be deployed; how well deployed resources will perform; how long cases take; how other variables, such as traffic, day or time of the week and weather affect incident patterns and responses and many more related questions.

Any robust deployment optimization using AI will also require plans that will work well with multiple incident scenarios and the overall objective must be to minimalize the failed incident response. Human controllers will need to carry out ongoing performance evaluation of these systems. This can be done by comparing deployments designed using generated incidents tested on actual data. Ultimately, evaluations will help improve the systems, since the machine learning model builds off past successes and failures.

CloudPrisma experts brings Smartness & Intelligence to the Resource Mobilization techniques adopted by Law Enforcement Agencies.