Вы находитесь на странице: 1из 13

Bid No (for Home Office

use only)

POLICE INNOVATION FUND 2015/16


Section 1 About the activity or project
1.1 Bid title
A Safer, Predictive City Using data feeds to proactively identify and alert officers to
threat, risk and harm.

1.2 Police forces and/or other partners involved

West Midlands Police


DVLA
Central Motorway Policing Group (CMPG) including Staffs Police
Highways Agency
Centro the West Midlands Transport Authority
Private Sector insurance companies
National Crime Agency (NCA)
Global Police Forces

1.3 Bid description


West Midlands Police has already lost 150m in funding over the last five years, however
the Force continues to investigate and adopt innovative methodology to continue delivering
policing services within our region. As with many other UK Police Force we are facing
budget reductions from both central government funding cuts, as well as limits to the
funding that can be gained from the precept. The funding challenge has been felt as
austerity previously but is expected to be felt more significantly over the coming years with
further comprehensive spending reviews.
However, the profile of crime and citizen demographics are also changing. Historic volume
crime such as burglary is dropping, and there is an increase in digital and cyber-crime in line
with improvements in technology, as well as an increasing focus on preventing vulnerability.
Within both the current austere environment, and expecting future significant funding
challenges, WMP are (in partnership with Accenture as the Forces Innovation & Integration
Partner) reviewing the policing services that they provide, how they provide them and the
level of service that customers and citizens should expect. This review will result in a new
operating model for WMP, that aims to move away from a reactive approach to policing, to a
more predictive and preventative model. This operating model will mean that the Force can
reduce demand before an event even requires a police interaction, as well as utilising their
resources in a more efficient way when they do have to respond. The move to a more preemptive and predictive operating model and culture is challenging, requiring not just budget,
but also a change in culture and strengthened partner relationships. The Home Office
Innovation Fund provides a route to support the Force with an initiative that otherwise the
Force would not have the budget or desire to commit to in a currently challenging financial
environment.

This bid is for a Safer City initiative. The sheer volume of data being captured currently
through CCTV and other cameras means that human analysis is simply not fast or thorough
enough to make best use of it; indeed, some estimate that up to 98 percent of CCTV
footage remains unseen by anyone. The Safer City initiative will utilise the many streams of
digital information and data that are recorded and shared across partners within a
geographic location (i.e. the West Midlands) to:

Proactively identify threat


Prevent risk and harm
Improve the utilisation of resources
Increase positive policing outcomes

In collaboration with partners within the West Midlands geography, analytics will be
permanently applied to integrated digital data streams to make the West Midlands a safer
place for people to live and work.
This will be achieved through:

Increasing the speed through which wanted or high risk individuals are found:
Automatic number plate recognition data will identify and alert officers when a watchlisted vehicle is picked up through CCTV cameras. Use CCTV to track patterns of
movement for known wanted offenders and identify locations where they can be
intercepted.
Disrupting criminal activity: Use convoy analysis on number plate recognition data
to identify vehicles used by organised criminal gangs, and keyless car theft. Analyse
movement for known drugs gangs to identify the best time to intercept, search and
disrupt their activity. Use social media monitoring to prevent disruption, and
automatically flag incidents.
Protecting vulnerable citizens: CCTV data from public transport providers (such as
Centro), shopping centres and public areas (e.g. libraries) will alert officers when a
missing individuals face is recognised. Number plate recognition data will allow
officers to track and monitor vehicles travelling together or patterns to vehicle
locations and movements, so that officers can identify and disrupt child sexual
exploitation.
Enhance shared intelligence: Vehicle hot-lists and facial recognition hits will be
linked to Force and partner intelligence systems so that a vehicle or persons
movements can form part of intelligence briefings and analysis. Real-time CCTV
analytics will alert officers and other intelligence organisations when a crowd is
forming and individuals are displaying aggressive facial triggers.
Increase speed to response: Response assets will be automatically scheduled and
deployed when the underlying analytics triggers an alert, so that officers and
emergency service partners can respond to and attend incidents more quickly. Realtime traffic monitoring will indicate the fastest route to an incident.
Support cross-partnership working: Identify cars with no insurance and link to
gang activity. Support multi-agency operations in locating wanted individuals across
geographic borders, e.g. with Europol and Interpol.

This is an innovative approach as analytics are not just focused on tactical crimes or issues,
but data stream are integrated to take a total place approach to understanding where a
threat may and either preventing or pre-empting the response required. A similar initiative
has been undertaken by the Singapore government, who through their Safer City initiative
have undertaken a pilot within two areas of the city to integrate advanced analytic
capabilities into the existing video monitoring systems,

to increase situational awareness, streamline operations and enhance the response times
of city authorities to public safety incidents. This has proven innovation in crowd
management analytics, with a greater than 80 percent accuracy using just five video
cameras in one of Singapores busiest metro interchange stations.
In addition, a crowd simulation model was used to predict crowd movement through this
station during the 2013 Formula 1 Grand Prix in Singapore. Other benefits gained include
increased operational support, anomaly detection through social media monitoring and
analytics and real-time decision making.
This will also result in cashable and efficiency savings for WMP and partners. Demand will
be reduced and intelligence automated which will result in staff savings in intelligence and
current teams that manually view CCTV footage, as well as staff savings in dispatch.
Cashable savings will be made from an improved utilisation of fleet which will reduce fleet
maintenance and running costs (e.g. reduced fuel cost as a result of reduced miles travelled
and reduced engine idle time).
Efficiency savings will be made as a result of improving officer utilisation and officers being
deployed to incidents with the highest chance of a positive outcome such as arrest or
penalty notice.
Non-financial benefits will include increased citizen satisfaction, increased positive justice
outcomes (e.g. increased early pleas, increased sanction rates), improved performance
(e.g. sanction detection rates, response to calls for service, reduction in total recorded
crime).
From a technical perspective, two years of historical data will be mined to establish a set of
rules (known as decision trees) that know what situations require an alert to be triggered,
and how the Force should respond to that alert (e.g. take no action, watch, or deploy a
response). Common data exchange standards will need to be established on the data
formats e.g. video using H264, as well as establishing IP/digital video management
solutions as opposed to analogue which limits the potential for better/faster streaming and
also better video quality for video analytics to be enabled. The infrastructure for the
networking technologies to transfer and share information/data will also need to be
established. Terminals must be then enabled to access and use the analytics, as well as
cloud storage for the data that is mined and analysed.

1.4 Key policy areas


Please check the box if this bid includes any element from these key policy areas. This is to
assist the sorting and scoring process only, and will not influence the decision on what bids
to award.
Blue Light Integration
Mobile IT
Body Worn Video
ICT change

Shared Services
Other (please specify)
______Big data, analytics, cross-partnership working___

1.5 Key objectives


The key objectives of this project will be to:
1. Review data and establish a set of decision trees that will trigger a specific
Force or partner response. For example dispatching a priority 2 response when a
watch-listed vehicle is picked up, or an unattended bag is pictured in a shopping
centre. We will use our relationship with Cambridge University to test and validate
these decision trees against current evidence-based policing trials. At the completion
of this first phase we will be able to share the analytic logic and decision trees with
other Forces who can utilise them for their own analytics initiatives.
2. Developing the policy and processes to support this from a business / operational
perspective, assessing and considering possible approaches to integration and
ensuring the bid fits in with the target operating model.
3. Undertake the necessary technology change that will be required to create an
integrated digital feed, and utilise the analytics tools and technology required.
4. Establish a capability within the Force to maintain the technology and analytics
tools, as well as understand and update the logic behind the decision trees.
5. Test the safer city initiative, including the logic paths, triggers, alerts and automatic
deployment processes in a single geographic area. We have had early
engagement with and commitment from Solihull Borough Council and Centro (the
West Midlands Transport Authority) to run this first test in Solihull. This test will allow
us to refine the logic and reduce errors prior to a full roll-out.
6. Roll-out the safer city initiative across the West Midlands geography, so that all
data feeds are integrated, and alerts and deployment of resources are centrally
managed, in line with the revised West Midlands Police 2020 operating model.

1.6 Key deliverables


The current partnership with Accenture allows WMP to procure the scope of this project
from Accenture and Accentures supply chain, providing that Accenture and any partners
can prove value for money and technical ability, without the need for a long procurement
exercise. Should the Home Office grant innovation funding in early March, this project would
commence in early April 2015 at the start of the new financial year.
Phase 1- April July 2015
Mine historical data to identify a set of rules and decision trees to trigger a specific response
in certain situations. For example what level of automatic alert should be provided and what
resource deployed when a wanted vehicle is picked up on CCTV.
The deliverable will be a set of decision trees and supporting processes, based on historic
data mining and evidence-based policing studies. This deliverable will be applicable to other
Forces who can use the decision trees and logic to prevent and pre-empt demand in their
own geographies.
Phase 2 August November 2015
Establish the infrastructure and software required to integrate the data sets and existing
security devices such as CCTV, sensors and detectors across organisations and apply the
analytics techniques.
Deliverable: Integrated data set, analytics tools and workflows, review of information
security controls to be implemented and reviewed to assess the impact of the bid within the
target operating model.

Phase 3 October November 2015


People within WMP, local authorities, transport agencies and emergency services trained
and capable to understand and update decision trees, and apply changes to the supporting
technology and workflow.

Deliverable: A training programme complete. A new capability will be embedded within


WMP that will allow other Forces to learn from and replicate the analytics approaches in
their own geographies.
Phase 4 November 2015 April 2016
Pilot of the safer city initiative within Solihull. Real-time use of analytics to deploy resources.
Integrated analytic feeds to provide intelligence and inform investigations and operations.
Deliverable: Safer City initiative live in Solihull. On-going monitoring and review of analytic
performance and error rates. Monitoring and review of accuracy of data feeds.

Phase 5 October November 2016


Revision of triggers, alerts, workflow and analytics techniques following lessons learned in
Phase 4. Roll out of the safer city initiative across the full West Midlands geography, based
on assessment of which LPU areas would be cost effective for implementation.
Deliverable: Safer City initiative live across the West Midlands. On-going maintenance and
support of analytics and underlying technology. A WMP-led capability that other Forces can
utilise to feed in their own data sets to take advantage of the integrated intelligence,
automated workflow and deployment.
1.7 Savings and benefits

Cashable savings
Efficiency savings

2016/17
1,400,000
60,000

2017/18
1,800,000
100,000

2018/19
1,800,000
100,000

Please also provide (where applicable) details of savings that are expected to accrue to
other public sector organisations as a result of the proposed activity/project.

Cashable savings will be as a result of:


Reduction in demand for manual intelligence analysis and review
Automation of CCTV triggering workflow
Reduction in call handling and dispatch staff effort
Improved utilisation of fleet which will reduce fleet maintenance and running costs
(e.g. reduced fuel cost as a result of reduced miles travelled and reduced engine idle
time)
Efficiency savings will be made as a result of:
Reduction in covert surveillance teams
Increased utilisation of response teams (proactive response during times of low
demand, e.g. CMPG)

1.8 Risks and mitigations


Risk: Could lead to a large number of false positives (leading to additional workload) or
missed matches (leading to reduced confidence in technology).
Impact: High
Likelihood: Low
Mitigation: Pilot decision trees and workflow on a set of test data to test triggers and
matches. Consider a phased roll-out of analytic capability on a reduced volume of data sets
to determine real-time errors and false positives.
Risk: Poor data quality results in additional cleansing effort required prior to implementation
of the COE
Impact: High
Likelihood: Medium
Mitigation: Assess quality of existing data sources and data warehouses. Commence data
cleansing as part of the existing portfolio of change within the Force.
Risk: Data required for analytics purposes is held by partners who do not wish to, or
cannot, share that data.
Impact: High
Likelihood: Medium
Mitigation: Engage specific partners in the creation of the analytics strategy to identify
available data sets and sharing approach, e.g. NCA, HMRC, DWP, education, local
authorities.
Risk: Citizens object to the data streams being used for crime prevention activities and
analysis.
Impact: High
Likelihood: High
Mitigation: Automating video capture enhances data privacy: specifically large-scale data
collection protects citizens anonymityby assessing patterns, not people. For the majority
of digital analytics applications, the data is used for counting purposes only, while special
cameras are required, beyond regular CCTV, for face recognition. Robust privacy laws and
security obligations mean that anonymised data is the only practical answer for most video
analytics applications.

1.9 Critical Success Factors


Data Quality It is expected that data will require cleansing so that it can be utilised in the
recommended analytics techniques, as well as being accurate inputs. Assess data quality
and plan cleansing activities into the schedule.
Data Integration Data sets must be integrated to avoid the effort and cost to perform the
analytics outweighing the expected benefit. Assess upfront in Phase 1 the data sets likely
and the complexities of integration. Continue the WMP2020 blueprint activities to define
data integration standards and data storage options.
Insight Leads to Action Analytics can only indicate where assets should be focused for
the optimum impact based on historical patterns: any insight gleaned from analytics must be
acted upon to achieve the benefits of the project.

Section 2 Total Costs


Please insert details of the overall costs associated with this bid. This should cover all
projected expenditure on the project/activity, regardless of funding source, and be profiled
against financial years (April to March).
Please insert:
A figure for the total overall costs for each financial year
Details of how this is broken down across your planned activities and whether the
spend is resource or capital.
2.1 Overall costs in 2015/16
Total overall costs:
Activity
Phase 1: Strategy and Decision Trees
Phase 2: Technology and Infrastructure
Cloud storage (IL3) annual cost
Analytics Tools & Licenses
Networking Infrastructure
Phase 3: Training and Capability Uplift
Access Terminals
Phase 4: Solihull Pilot
Project Management
Service Support

2.970m
Cost
800,000
600,000
80,000
100,000
40,000
500,000
20,000
500,000
250,000
80,000

Capital/Resource
Resource
Resource
Capital
Capital
Capital
Resource
Capital
Resource
Resource
Resource

2.2 Overall costs in 2016/17


Total overall costs:
Activity
Cloud storage (IL3) annual cost
Phase 5: Full Roll Out
Access Terminals
Service Support

2.72m
Cost
80,000
2,500,000
60,000
80,000

Capital/Resource
Resource
Resource
Capital
Resource

Section 3 Funding Required


Please note that all funding needs to be profiled against financial years (April to March).
Please insert:
A figure for the total amount requested for each financial year
The percentage of your total costs this represents
Details of how this is broken down across your planned activities and whether the
spend is resource or capital.
3.1 Funding requested for 2015/16
Total Police Innovation Funding requested:
Percentage of overall costs (as in section 2.1)
Activity
Phase 1: Strategy and Decision Trees
Phase 2: Technology and Infrastructure
Cloud storage (IL3) annual cost
Analytics Tools & Licenses
Networking Infrastructure
Phase 3: Training and Capability Uplift
Access Terminals
Phase 4: Solihull Pilot
Project Management
Service Support
3.2 Funding requested for 2016/17
Total Police Innovation Funding requested:
Percentage of overall costs (as in section 2.2)
Activity
Cloud storage (IL3) annual cost
Phase 5: Full Roll Out
Access Terminals
Service Support

1.753m
59%
Contribution to
funding
472,000.00
354,000.00
47,200.00
59,000.00
23,600.00
295,000.00
11,800.00
295,000.00
147,500.00
47,200.00

1.605m
59%
Contribution to
funding
47,200.00
1,475,000.00
35,400.00
47,200.00

Capital/Resource
Capital/Resource
Resource
Resource
Resource
Capital
Capital
Resource
Capital
Resource
Resource

Capital/Resource
Capital
Resource
Capital
Resource

Section 4 How the activity or project meets Innovation Fund criteria


Please provide a brief summary of how your bid addresses each of the Innovation Funds
bidding criteria (see separate guidance). Maximum 300 words per criterion.
A. Transform policing through innovation
It is recognised by WMP that the operating model for policing has to change in order to
respond to the reducing budgets and changes to crime and demographic profiles. The act
of policing is also changing and becoming more digital, therefore digital capabilities need to
be built and exploited. The safer city initiative is fundamentally an enabler to changing the
operating model of policing in the UK, through a more predictive and pre-emptive approach
to reducing crime, whilst breaking historic silos between different public service
organisations.
The safer city project will for the first time allow WMP and public service partners to use
intelligent, predictive, repeatable analytics to predict and pre-empt threat, risk and harm.
Innovative technologies and analytics techniques will be used on data sets that have never
previously been valued as a source of intelligence or trigger for process automation. It will
utilise a previously untapped data source to identify where fleet and response teams can be
scheduled and deployed to intercept wanted individuals, reduce and prevent crime. This
approach to predictive analytics is not currently being explored by another other UK Force,
and builds on other smaller predictive analytics pilots.
The approach to focus on a geographic area in the total place mentality is also innovative,
breaking across many of the existing silos between public service organisations and data
owners.
By understanding which situations will have a higher likelihood of a positive resolution,
officers (and fleet) utilisation will be improved, with response officers being targeted in the
right place at the right time. Furthermore, the ease and speed with which insight can be
gained from automatic alerts and tasking will free up officer and staff time which would
currently be spent on manual analysis, interpretation and reporting of intelligence.

B. Enhance collaboration
The fundamental success of the safer city project is dependent on collaboration across
partners to establish a united security system and data sets. This project will build on the
relationships and partnership facilitation that is already underway between the Force and
partners as a result of the WMP2020 vision, providing a project with tangible benefits to
further build and strengthen cross-partner collaboration.
The project will also further cement joined-up aims and objectives across partners, focusing
on fighting crime and reduce threat, risk and harm to the citizens and businesses of the
West Midlands, rather than to benefit only the Force.
A richer data set will be created that partner organisations can utilise for additional analytics.
The safer city project will result in a collection of reusable assets at different levels that will
be shared with other Forces should they wish to establish a similar safer city approach.
These will include the triggers and decision trees for the analytics capabilities, as well as the

technical architectures that describe how the data sets and infrastructure are linked together
in a secure environment.
Once the full safer city roll-out is complete, the initiative will be able to scale and be
commoditised for other Forces to plug their data feeds in to the initiative and take advantage
of the capabilities and infrastructure.
Early engagement with Solihull Borough Council has resulted in their commitment to this
project and Solihull being the test-bed for the first roll-out of the analytics capabilities.
Similarly Centro (the West Midlands transport executive across bus, tram and train) has
also committed to participating in this project should innovation funding be granted.

C. Deliver efficiencies
The safer city project will result in cashable and efficiency savings for WMP.
Cashable savings will be as a result of:
Reduction in demand for manual intelligence analysis and review
Automation of CCTV triggering workflow
Reduction in call handling and dispatch staff effort
Improved utilisation of fleet which will reduce fleet maintenance and running costs
(e.g. reduced fuel cost as a result of reduced miles travelled and reduced engine idle
time)
Efficiency savings will be made as a result of:
Reduction in covert surveillance teams
Increased utilisation of response teams (proactive response during times of low
demand)
Non-financial benefits will include:
Increased citizen satisfaction and public perception of the Force
Increased positive justice outcomes (e.g. increased early pleas, increased sanction
rates)
Improved performance (e.g. sanction detection rates, response to calls for service,
reduction in total recorded crime)
Improved intelligence picture and automatic alerts (e.g. linked people, locations and
events)
Increase in speed to identify wanted vehicles and individuals
Increase in recapture of stolen vehicles
Reduction in overall WMP demand and calls for service

D. Be locally owned and monitored


The safer city project will be managed and run by the West Midlands Police, with
engagement and input from other partner organisations at each phase of the project.
A project team and governance board will be established, which will monitor monthly

progress against plan, including budget, review of data quality, analysis and integration. The
WM OPCC will be represented on this board, with a public report being made available
three months after Phase 4 completes, to provide an update on the benefits and success of
the project (note there must be a lag on any public reporting and public information
censored to ensure efficacy of the analytics techniques).
This bid is for the Innovation Fund to match 40% of the costs to implement the safer city
projects: the remaining 60% will be funded by West Midlands Police.
The success of the project implementation itself will be assessed against meeting project
deliverables without additional expenditure. The success of the initiative will be evaluated
and measured through the:
Reduction in total recorded crime
Reduction in budget annually
Public satisfaction scores
Number of sanctions
Detection rate
Percentage of successful stops following automated workflow deployment
Number of detained persons brought into custody

Section 5 Police and Crime Commissioner sign-off


Please confirm that this bid is signed off by the Police and Crime Commissioner for the lead
force area and (if applicable) for all other forces named at section 1.2 of this bid. All bids
should also confirm sign-off by the OPCC Chief Finance Officer for the lead force area.
Where this bid involves collaboration with other public sector or third sector organisations
this section should also confirm sign-off by the relevant responsible party(ies)

Section 6 Contact details


6.1 Lead OPCC Chief Finance Officer details
Name:
Phone:
Email:
6.2 OPCC/force project contact details
Name:
Role:
Phone:
Email:

Вам также может понравиться