The problem

The visitor is uncertain that their booking has been made, and HMPPS is unsure that the visitor is who they say they are.

The current service involves booking requests from visitors, and the entire submission and approval workflows rely on manual data matching and scheduling by the prison officers.

Some visitors do not attend or are turned away as they mistake confirmation of their booking request for confirmation of a booking slot.
Some visitor data is incomplete and has errors and there are multiple profiles created in NOMIS (one of the legacy systems used for the management of offenders in prisons) for a single visitor as a result of human error.

This means that the visiting process is inefficient; visit centres have empty seats, prisoners do not receive their full entitlement, and intelligence teams are lacking the information they need.

Empathise

Families that are new to the prison system need help with the visit process, from sentencing in court through the gate to meeting their loved ones on the visit day.

The team identified key user groups and discovered their issues when booking, preparing and attending the visit. 

Families were confused and stressed after the sentence since it was challenging to locate the prisoner. There is a service called 'Find a Prisoner,' but it is useless unless the family has the prisoner's ID number, which is issued to the prisoner upon arrival at the prison.

Visitors were not given clear information about the rules at the point of booking. This caused issues ‘through the gate’ as those rules (i.e.clothing, lockers, money, cigarettes, searches etc) were flagged too late in their journey.

Prisoners were relied on to share visit information with visitors but couldn't access up-to-date changes in visit processes. They were frustrated by the lack of transparency and wanted to know who is visiting in advance and be able to turn visitors away if they don't want to see them.

The current booking ‘request’ process goes through too many people, steps and systems - not real-time or digital by default. Human error leads to duplications and incomplete visitor data.  

HMPPS can not compile accurate visitor data. ID processes differ
in each prison, even if in the same category. Security was compromised as there was no way to monitor visitor behaviour over time across the estate.

End-to-end user journey

User groups, needs and pains

Service blueprint as-is

Policy map as-is

Define

3 key areas to look into:
Information, automation and identification.

The service blueprints were mapped to visualise how the current service works and identified significant pain points in user flows, policies and regulations. We defined 3 themes to test. 

We wanted to build a system that meets visitors' informational needs in order to reduce their fear and anxiety about prison visits, provides instant booking confirmation and consistent user experiences for visitors and prison officers, and collects accurate and complete visitor data to improve prison security.

How Might We...

Riskest assumption 1

If we create a standardised booking system, the majority of prisons will be able to implement and operate it effectively.

The legacy system has been used by more than 170 booking agents and prison officers across 200 prison estates in the UK. Each prison has more than one visit location in the database to assign visit slots to prisoners. These locations are not always physical rooms but specific visit types that are classified and listed in NOMIS. 

Visit slots are allocated to prisoners based on prisoner category, visit type, and capacity per visit type. The ability to determine if a prisoner comes under a certain visit type, where different visit periods and capacities may apply, is extremely important.

To standardise the booking procedure, we needed a better understanding of different operation models per prison and validation of what works for the majority of institutions.

We also discovered that inconsistencies in booking schedules have an impact on user experience because visitors aren't always aware of what adjustments are being made. This can also lead to a high volume of phone calls.

Riskest assumption 2

An automated system will improve operational efficiencies while saving users time and effort.

Before Covid, around 20,000 visits were booked weekly, with a 50% digital uptake for the online booking system. For visitors, it can take up to 3 days to receive confirmation of a booking online.

In order to desgin real-time booking confirmation workflow there were a number of things we needed to investigate further, such as which parts of the booking process we could or could not automate, and which types of visits could be automated and handled centrally.

We also discovered that some prisons took different approaches to operate multiple visit locations in the system. For example, one prison has 2 booking teams that are working across the same process but for different channels. Team 1 handles counter and telephone booking requests and is extremely busy with no process lead time. Team 2 deals with online, email (274 backlogs) and telephone requests. 

This may affect user experience and result in a high rejection rate since online booking requests are never given priority. In order to establish a centralised approach, we would need to support prisons in streamlining their visit booking procedures and teams.

Riskest assumption 3

We can build a visitor database that will help us accurately identify a visitor on arrival.

The current visitor database is prone to visitor identity fraud and misuse. The system must be able to distinguish duplicate or incomplete visitor profiles and maintain only the correct information.

The visitor registration process is time-consuming. Adding visitor information to NOMIS can take up to 3 weeks. Before data is recorded for visitors to book a visit, it goes through numerous manual checks with central and local security policies and restrictions. Meanwhile, there is no communication with visitors or prisoners about their registration status, leaving them in the dark.

Based on the assumption that all prison estates implement a central visitor registry, we hypothesised that only validated visitor IDs would be recorded in a central registry in NOMIS, which would link to inmates who had been visited by the identified individuals.

We also assumed that we might be able to link those visitor IDs to various credential sets, such as online multi-factor, biometrics, or documentary evidence, to fulfil the security needs of each prison. However, we must ensure that all visit interactions check the visitor's credentials against the visitor registration.

So the first thing to define was what minimum data we need to catch from visitors and how those can be validated as quickly and easily as possible. We have done a few workshops with people from Security and came up with some potential concepts for the new visitor registration service.

User journey mapping exercise

Idea sketches

Explore

The service to-be:
digital service that enables visitors easily book, cancel and amend a prison visit.

First of all, we designed a few user flows to test with family and friends of prisoners. We focused on the visitor registration journey, including ID validation, and the prison visit booking journey. And then we explored what feasible technologies and processes we can employ to support the registration and booking services for visitors.

To make the standard prison visit booking procedure aiming to perform automatic booking confirmation for the majority of prison estates we designed a number of scheduling and timetabling visit slots work flows, which can be test with typical cat D prison to determin whether it meets the purpose under different prison's restrictions and regulations.

Shaping user journeys

Service blueprint to-be

Prototype for prison visit booking service

Copyright ©SingBlue 2019. All rights reserved.