Predictive Modeling is the process of using known results to create, process, and validate a model that can be used to forecast future outcomes.  Deccan runs a thorough analysis of the client’s historical data and then leverages that historical performance to predict how the agency will perform under variable future conditions.   

Predictive Modeling is important because it allows clients to demonstrate or simulate their performance before future conditions occur.  For example, if a developer plans to build 1000 new homes in a particular part of the community, they can analyze how their current resources will respond over the proposed road network and to the increased call volume the new development will bring. Predictive Modeling applications graphically depict the impact of growth on response times and the ability to meet community emergency services needs. Additionally, analysis can be run to ensure that any new stations offered as part of the development are placed in a way to benefit the entire community. 

Contrary to modeling community growth, analysis can also be run to demonstrate the operational impacts of funding/staffing reductions. Response capabilities can be simulated without the resources in question, visually demonstrating what the impacts of funding reductions will be, along with who will be impacted by those reductions. 

While predictive analytics describes a set of conditions in the future based on historical analysis, prescriptive analytics provides recommendations on the appropriate action to take for optimal outcomes.  Deccan clients benefit from prescriptive analytics through real-time analysis of their operational deployment, along with recommendations on where to locate remaining resources for optimal coverage at any given time.  Such prescriptive analysis monitors operational deployment 24 hours a day, seven days a week, to ensure the best deployment coverage.  This is critically important as it can shave minutes off response times to emergencies.

Predictive and prescriptive analytics leverage historical response data to build an algorithm that can simulate agency response based on a given set of conditions.  The algorithm considers response times, road network, traffic speeds, historical traffic conditions, and historical unit availability to create simulations based on client queries.

Operational Intelligence is leveraging data and analytics in a way that generates real-time and relevant information that guides decisions in the emergency services environment.

Run cards or “run orders” are a sequence of resources designated to respond to a particular type of emergency in a specific geographic space.  The run cards are loaded into the dispatch software (CAD) for public safety dispatch entities.

Manually updating a set of run cards can be a very tedious process.  It requires staff to manually consider each set of call types for each geographic response zone listed in the CAD.  Manual configuration of run cards can take weeks, if not months, to update and are subject to human error.  Deccan clients use analytics to automate this process by verifying street networks, historical traffic conditions and predicted response times to automate the selection of each resource for each call type.  This analysis saves tremendous time for staff and eliminates the opportunity for human error.

Predictive Modeling can help create incident risk predictions for communities and target hazards based on various building and demographic characteristics and historical trends. These predictions can then be used by departments to prioritize their community risk reduction efforts, making sure that they’re making optimal use of their limited resources. The battle for survival begins long before ignition. Predictive modeling allows public safety agencies to reduce response times by optimizing resource placement before the emergency occurs.

Estimated Time Back (ETB) is a forecast for when a resource will be available to provide coverage. ETB, along with call volume predictions, impacts recommendations for re-positioning resources by minimizing then when a resource is expected to be available soon and encouraging them when a resource is expected to be busy for lengthy periods, depending on the expected call volume.

A scoresheet could include, response, vehicle, or incident criteria. Response criteria show response performance to incidents and are usually measured in seconds or as a percentage when measuring compliance to a time target. Vehicle, or unit, criteria are metrics related to the performance of vehicles, such as Unit Hour Utilization (UHU). Incident criteria are metrics related to incidents, such as incident volume.

A forecast for when the resource will be available to provide coverage

Re-positioning of resources to maximize coverage

A collection of activities to prevent emergency incidents from occurring within communities

AKA Public Safety Answering Point (PSAP), a location where emergency calls are received and resources are dispatched and monitored

Pre-set plans to reposition resources based on predefined statuses of the system

Probability of having enough resources to respond to incidents based on historical trends

An area that is likely to have an incident with a lengthy response time due to unavailable nearby resources

An area that is served within a specific response time by the same number and location of resources

A metric that incorporates the probability of having an incident with a lengthy response time due to unavailable nearby resources across the entire jurisdiction

Historical data on incidents, including their time, location and type

Historical scenarios are constructs that enable a Deccan user to perform analyses on operational data relating to a specific time-period. Different historical scenarios can be created for incident and response data relating to different time-periods.

Incidents refer to emergency calls that are received, responded to by an emergency vehicle, and catalogued in CAD.

Street Network is a set of street segments within the area being analyzed. These street segments have information regarding addresses, speed limits, elevation, and direction.

A response plan refers to the types and number of capabilities that are dispatched to a call of a particular type.

ITGs, or Incident Type Groupings, are broader segments of types of incidents. By categorizing incident types along with other incident types that are like each other, we can do analysis more effectively.

A scoresheet contains several response, vehicle, and/or incident criteria along with their scores for their time-period being analyzed.

Message boards are platforms for relevant information to be communicated and displayed to multiple stakeholders.