Service Definition

FUJITSU G-Cloud Services { Insert Service Name }

Automated Inspections
through Artificial
Intelligence

G-Cloud 11

Page 1 of 12 ©Copyright FUJITSU 2019

Table of Contents
1. Automated Inspections through Artificial Intelligence ...............................................................................3
2. Service Delivery Approach..............................................................................................................................5
3. Service Deliverables .......................................................................................................................................7
4. Pricing .........................................................................................................................................................8
5. Commercial .....................................................................................................................................................9

List of Figures
Figure 1: automation of inspections through exceptional accuracy levels ................................................................... 3
Figure 2: Service Delivery Approach ............................................................................................................................. 5

List of Tables
Table 1: Day Rates for the Provision of Fujitsu Professional Services Resource ............................................................. 8
Table 2: Definitions ................................................................................................................................................... 11
Table 3: Table Action Caption ............................................................................................ Error! Bookmark not defined.

Page 2 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

1. Automated Inspections through Artificial Intelligence

Today many organizations still depend on skilled workers to undertake repetitive visual inspection across
many critical processes, for example rail and roads in transportation, and video covering vast areas
captured via drone. This adds significant cost and inefficient to these processes, in a period where price
pressures are ever increasing. Current processes are also susceptible to human error, risking missing
critical issues resulting in potential reputational damage. Fujitsu has proven how Artificial Intelligence (AI)
can be applied to help organizations respond to these challenges, enabling automation of such
inspections through exceptional accuracy levels.

Figure 1: automation of inspections through exceptional accuracy levels

This service is based on applying the FUJITSU Advanced Image Recognition (F|AIR) solution, proven to
revolutionize any operation involving a visual inspection to assess the condition of an asset or material.
Through applying Artificial Intelligence Deep Learning, it delivers greater business insight and flexibility
than was possible with previous automated visual inspection capabilities. It enables this through:
 Training from a sample of images which contain examples of the defects to be identified –
enabling fast implementation and refinement to support varied and changing inspection
requirements
 Managing ambiguity in what it is looking for – the solution is not looking for something exact in
an image, but something similar to what it has been trained to recognize. This is important for
enabling coverage across a wide range of defect types and high levels of accuracy.
Any type of image is supported, for example Ultra-Sound, Infrared and Thermal; thus enabling use of the
most effective technique for each specific type of inspection.
Training of the engine involves ingesting images containing features of interest. By comparing these with
data containing no features, Fujitsu’s solution generates a recognition model that is able to identify the
difference between the two. This model is then applied to automatically highlight areas of interest across
images of all subsequent inspections. The user interface presents its findings in a format most suitable for
the image content being analysed.
The image recognition platform is designed to adapt to a customer’s current environment. It can be
deployed on premise or in the cloud, it offers an API to enable integration with existing manufacturing
systems, and can easily be extended to accommodate specific requirements. An audit is captured across
all quality inspection findings and decisions, and this is compiled into a PDF report which also contains
images of where the defects had occurred.
A key benefit of Fujitsu Advanced Image Recognition is how quickly it can be proven and then
implemented, ensuring a low risk engagement model and fast time to value. Fujitsu Advanced image
recognition can be applied to deliver significant business across many public sector services, for example
inspection of:

Page 3 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

 Rail through analysis of footage captured from cameras installed onto the trains
 Roads through analysis of footage captured from cameras installed onto vehicles
 Vegetation through analysis of footage captured across large areas of land via drone
 Building and other large assets via analysis of drone footage
 Surveillance camera footage to identity suspicious patters and behaviors
This is a platform that combines ready-to-use tools with an ability to adapt and extend to the end-users
needs. This means that organisations are finally able to apply AI to create real transformation, with lower
risk and demonstrable outcomes.

Page 4 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

2. Service Delivery Approach

A key benefit of the F|AIR solution is that it can be trialled very quickly to prove potential business value. It
is also a framework which then enables fast implementation, providing all the capabilities required to
apply AI across critical inspection processes. The high level phases of the service delivery approach are
described in the diagram below.

Figure 2: Service Delivery Approach

Feasibility Study
The first step is a brief Feasibility study exercise to confirm that F|AIR is able to work with the inspection
content and has potential to deliver required business outcomes. This involved obtaining a sample
dataset against which the F|AIR team will remotely undertake the model training to identify optimal
configuration. A trial will then be executed using a test data set (witnessed by the client) where success
will be measured based on accuracy levels and projected time savings

Proof of Value
Objective is this step is to confirm that applying F|AIR will achieve defined business objectives. Work
includes confirming the business case at a more detailed level, typically through workshops with
operational and inspection teams, typically quantifying benefits such as:
 Time and efficiency gains
 Improved inspection accuracy through higher accuracy levels and reduced human error
 Freeing up staff to focus on great value add activities
 Reduced staff training overheads
 Financial savings resulting from the above

Page 5 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

Minimum Viable Product (MVP) / Full Implementation / Scale Out

Phases for rollout of the solution:
 MVP: initial deployment of the core solution. Enables delivery of benefits of the solution whist
broader less critical aspects are delivered
 FULL IMPLEMENTAION: deployment of the full solution, for example enhanced reporting and
integration with operational systems
 SCALE-OUT: Broader rollout of the solution across further inspections processes and locations
where applicable
Through each of these phases, Fujitsu will provide support of the F|AIR application, as well any data
capture hardware where this is in scope.

Page 6 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

3. Service Deliverables
Deliverables from POV Phase
 Report containing:
o Results from PoV trials
o Proposed rollout solution design and approach based on PoC findings
o Recommended next steps
 Outline Business Case
 Presentation of content from above and workshop to agree how to proceed

Deliverables from MVP / Implementation

 Recognition models trained to identify specific inspection content
 Configured instance of F|AIR tailored to specific requirements, including integration with any data
capture hardware or operational systems, deployed either on cloud or on-premise
 Ongoing service management and operation of the F|AIR software

Page 7 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

4. Pricing
Fujitsu offers a number of commercial models for delivery of projects such as fixed price, Time and
Materials and Consumption Based “As a Service” models. Fujitsu welcomes the opportunity to discuss
specific requirements and propose the best suitable commercial model for the engagement.
Our rates are based on the Fujitsu’s published GCloud SFIA rate card.

ROLE PRICE (£)

PoV £20-50k depending on scope
Production Solution £2-10k per month depending on
scope

Table 1: Day Rates for the Provision of Fujitsu Professional Services Resource

Page 8 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

5. Commercial
Ordering and invoicing Process
 We will commence delivery of the service on receipt of a purchase order
 Fujitsu shall invoice for the Service plus any applicable taxes after delivery of all our products
defined in the agreed Terms of reference. Any additional expenses shall be invoiced in the month
after they have been incurred
 The customer will pay the invoiced amount in full within thirty (30) days of the date of each
invoice (the “Due Date”)
 If the customer has not paid Fujitsu an invoice by its Due Date, Fujitsu may charge interest on a
day-to-day basis from the due date at the rate of 4% per annum over the Barclays Bank Base Rate
until the payment is made in full
 The customer shall not be entitled to offset any sums owed by Fujitsu under any contract or
dispute between the customer and Fujitsu against any sums that the customer owes to Fujitsu
 When remitting payment, the customer will include the applicable Fujitsu invoice that the
payment applies to

Customer Termination
The Customer may terminate this Agreement [or any part thereof], at any time, for convenience by giving
not less than thirty (30) days written notice to Fujitsu.

Supplier Termination
Fujitsu may terminate this Agreement [or any part thereof], at any time, for convenience by giving not
less than thirty (30) days written notice to the Customer.

Consumer Responsibilities
Successful delivery of the Assessment Service is subject to the following dependencies upon the Customer:
 Appropriate communications and engagement within the Customer organisation
 Appointment of a single point of contact for the Assessment
 Any information available and required for the assessment will be shared with Fujitsu during (or
within 5 working days after) the Alignment Meeting.
 Timely provision of ongoing data as requested by Fujitsu
 Timely access to appropriate/suitable resources
 Co-ordination and timely scheduling (and facilities) for workshops (and meetings), ensuring
appropriate attendance and active participation

The Customer acknowledges that the timely and adequate compliance with the obligations above is
essential to the performance of the Fujitsu Assessment Service. Fujitsu will not be liable for any delay or
deficiency in providing the Service if such delay or deficiency results from the Customer’s failure to fulfil
these dependencies. Should a delay to the Service result from the Customer’s failure in relation to the
above dependencies, for example, if there is not full attendance at the scheduled workshops, Fujitsu shall
be entitled to amend the Service, Schedule and/or Charges with no liability and shall be entitled to charge
Customer for any cost incurred as a result.

Page 9 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

Should the Customer request (and Fujitsu agree) that Fujitsu consultants undertake any of the obligations described
above Fujitsu reserve the right to amend the Charges.

Service Constraints
N/A

Service Exclusions
N/A

Page 10 of 12 ©Copyright FUJITSU 2019

FUJITSU G-Cloud 11 Service - Automated Inspections through Artificial Intelligence

6. Definitions
Description Description
FAIR Fujitsu Advanced Image Recognition
POV Proof Of Value
MVP Minimum Viable Product

Table 2: Definitions

Page 11 of 12 ©Copyright FUJITSU 2019

About Fujitsu
Fujitsu has been working with the public sector for over 40 years and is a global IT company offering a
complete range of products, services and solutions. From looking after applications and protecting
data, to managing supercomputers around the world, we’re helping government and business
everywhere to become more innovative and efficient. Winners of Business in the Community’s (BITC)
Responsible Business of the Year 2015 and with a 5* rating in their Corporate Responsibility Index, we
are also proud to work alongside our charity partner MacMillan Cancer Support.

Subject to contract.
FUJITSU
© Copyright FUJITSU 2019. All rights reserved. FUJITSU and FUJITSU logo are trademarks of Fujitsu Limited registered in many
Phone: 0843 365 1549 jurisdictions worldwide. Other product, service and company names mentioned herein may be trademarks of Fujitsu or other
E-mail: government.frameworks@uk.fujitsu.com companies. This document is current as of the initial date of publication and subject to be changed by Fujitsu without notice.
This material is provided for information purposes only and Fujitsu assumes no liability related to its use. Subject to contract.
Website: uk.fujitsu.com Fujitsu endeavours to ensure that the information contained in this document is correct but, whilst every effort is made to
ensure the accuracy of such information, it accepts no liability for any loss (however caused) sustained as a result of any error or
omission in the same. No part of this document may be reproduced, stored or transmitted in any form without prior written
ID-3275/05-2016
permission of Fujitsu Services Ltd. Fujitsu Services Ltd endeavours to ensure that the information in this document is correct and
fairly stated, but does not accept liability for any errors or omissions.

uk.fujitsu.com
Page 12 of 12 ©Copyright FUJITSU 2019