Policy paper

Technology innovation in government survey

Published 20 August 2018

Preface

In 2017, the Director General of the Government Digital Service (GDS) commissioned a landscape review of technology innovation across government from Martin Smith, an independent contractor.

The objective of the survey was to understand current activity across government in what might be termed new or emerging technologies that are related to digital or information technologies. Loosely defined, these are new technologies that do not currently have a critical mass, but which may have the potential to disrupt industries or generate significant savings. Conversely, these may raise questions of ethics or have a structural impact on public services if employed (for example, some could lead to significant workforce shifts).

Specifically, we sought to understand:

• which specific emerging technologies departments and agencies are interested in and why
• how new technologies were being used to meet specific use cases
• how government organisations structure their experiments with new technology
• what innovation networks, communities and hubs exist around government
• broadly, what innovation capability exists within government

This research was carried out over the course of 2017. It is presented as a set of initial findings and should not be seen as an authoritative list, although it may form the basis of a register in the future.

It is limited to the digital function, so excludes technology innovation in, for example, military or healthcare hardware, although this distinction is becoming more blurred (in the case of autonomous vehicles, for example). It involved informal meetings with chief technology and digital officers, innovation heads, technology policy advisers and leaders from the Digital, Data and Technology profession (DDaT) in central departments, agencies and arms length bodies, as well as from within GDS. The data presented is not intended to be complete, but to provide a sense of current activity. This is particularly true for information relating to local authorities which should be seen as indicative only.

While this has been led from GDS, we have worked closely with the Department for Business, Energy and Industrial Strategy (BEIS), the leading department on industrial strategy and Department for Digital, Culture, Media and Sport (DCMS) in their leading role to:

• drive the ongoing success of the UK digital tech sector
• deliver a world-leading digital economy that works for everyone
• ensure the UK continues to be recognised as an attractive location for digital and tech sector businesses

The report that follows is based on Martin Smith’s original version, updated within GDS to account for factual changes since it was produced. The recommendations have also been updated to reflect current thinking.

1. Summary

GDS will lead work with departments to better coordinate, share best practice and drive technology innovation in government.

This report provides a review of the technology innovation landscape in government and provides some initial recommendations for further GDS work. It is published within the context of increasing demands on GDS from departments and other public bodies to do more to enable and support the innovative use of new emerging technologies - such as artificial intelligence, distributed ledgers and the internet of things - in government.

The Artificial Intelligence Sector Deal, published by BEIS and DCMS in March 2018, set out the government’s commitments on AI, within the context of the 2017 Budget commitment to increase research and development spend from 1.7% of GDP to 2.4% by 2027.

Technology innovation in government is therefore expected to grow rapidly in the coming years. GDS has recently set up a new Innovation team to enable and support public bodies to make best use of emerging technologies to:

• transform services
• increase productivity
• grow the nascent GovTech sector

This new GDS Innovation team includes the GovTech Catalyst that is leading work with departments on the £20 million GovTech Fund announced by the Chancellor in November 2017.

This survey is largely aimed at a public sector audience, though academics and government watchers may also find it helpful. It is published along with the release of the underlying data in the spirit of open innovation. It sets out what we have learnt to date about the technology innovation landscape across government and provides an initial roadmap of recommendations for the new GDS Innovation unit for 2018 to 2019.

Technology innovation is vital for the public sector. The current picture across central and local government shows lots of positive activity but also that much of it is uncoordinated and fragmented. GDS Innovation plans to help by sharing results, resources, best practice and working collaboratively on common blockers.

This survey is a starting point for GDS Innovation. During 2018-19 the team will work with other Cabinet Office functions, departments and the devolved administrations to develop a first ever innovation strategy for UK public services with a view to publication in spring 2019.

2. Introduction

In 2017, GDS’s Director General commissioned a survey of technology innovation in government. There is no single concrete definition for innovation and its use in the public sector. It means different things to different people. One definition (focused on Canada) categorises public sector innovation into 4 overlapping types: service, process, regulatory and policy. Technology as an enabler underpins them all, but technology innovation is a fifth type in its own right.

Table 1: The 5 types of public sector innovation

Innovation type	Description
Service innovation	Develop a new service concept or step change improvement to an existing service.
Process innovation	Rethink entire end-to-end processes that bring significant efficiency and productivity improvements.
Regulatory innovation	Support new business models and disruptive technologies through regulation and enforcement frameworks while protecting the public and stimulating the economy.
Policy innovation	Identify constituents’ needs and reduce development, testing and implementation times for new policies.
Technology innovation	Explore emerging and disruptive technologies or combine existing technologies in new ways to develop novel solutions and services.

2.1 How technology innovation relates to digital transformation

GDS published the Government Transformation Strategy 2017-2020 (GTS) in February 2017.

The GTS guides government in its delivery of world-class services. Technological innovation is part of the GTS and it includes examples of innovation from various transformation programmes. Keeping up with technological change and experimenting with new technologies is vital. It allows government to identify which technologies can assist the delivery of transformation programmes.

The government’s Industrial, Digital and Cybersecurity strategies also emphasise technological innovation. This is particularly the case during interactions with academia, industry and the market. The strategies guide government as it seeks to grow the digital economy, boost the growth of different industry sectors and keep national services and infrastructure secure.

2.2 Survey objectives

Other objectives of this technology innovation survey were to understand:

• what specific emerging technologies departments and agencies were interested in and why (and where possible to have an indicative view of what is happening in local authorities)
• how new technologies were being used to meet specific use cases
• how disruptive technologies such as artificial intelligence (AI) and distributed ledger technology (DLT) may impact future services and policies
• the innovation networks, communities and hubs around government
• innovation capability within government

2.3 What was done

The survey was conducted by Martin Smith - an independent contractor - throughout 2017. It involved informal meetings with IT directors, chief technology officers, innovation heads, technology policy advisers and leaders from the Digital, Data and Technology profession (DDaT) in central departments, agencies, arms length bodies and local authorities.

Data was collected from GOV.UK, external websites and discussions with experienced GDS staff from the Strategy and Engagement, Service Design and Assurance, National, International and Research, and Delivery and Support teams.

The scope of the survey included central government departments and agencies, regulatory bodies and local authorities. Views from industry and academic partners, think tanks and consultancies associated with government were gathered. The results are significantly more robust for central government than for local authorities, where any references are purely indicative.

This report was originally written by Martin Smith, though the framework for the recommendations and some of the detailed recommendations were provided by Graham Walker, the Deputy Director for GDS Innovation.

2.4 Related government reviews

In summer 2017, the Cabinet Office also commissioned an independent strategic market review of GovTech.

While administered by GDS, other government departments have been heavily involved: particularly DCMS, HM Treasury (HMT) and BEIS. BEIS is well placed to make a unique contribution based on its credibility and expertise as the business department with a particular ability to engage peers as fellow policy professionals to drive demand.

Two other technology innovation related government reports were published in 2017:

• Growing the artificial intelligence industry in the UK
• Delivering better outcomes for citizens: practical steps for unlocking public value

Growing the artificial intelligence industry in the UK urges government to lead on AI with a set of specific recommendations for GDS (see The impact of AI in government services under 3.3).

Delivering better outcomes for citizens: practical steps for unlocking public value states that government should support incremental and disruptive innovation. It should also drive efficiencies through technology and data to achieve marginal gains.

What is GovTech?

GovTech is a nascent but promising new industry sector where private sector startups and technology firms deliver innovative technology-based solutions (often using the latest technologies) to help solve public sector problems.

Other definitions from different industry sources and an overview of the ecosystem are in the appendix.

GDS managed the Cabinet Office GovTech review and identified common links with technology innovation research. This included existing government backed accelerators, incubators and catalysts, sources of innovation funding and how emerging technologies were being used by GovTech firms.

Working with HMT, BEIS, DCMS, Innovate UK and the devolved administrations, GDS established a cross-department GovTech programme.

In November 2017, a £20 million GovTech fund was announced in the Budget and Industrial Strategy. The fund and programme is overseen by a dedicated GovTech Catalyst team in GDS.

The programme funds competitions in the market for GovTech firms to solve departments’ and public bodies’ challenges. It also provides a means for GovTech firms to engage with government and grow the GovTech sector.

The first 5 public sector challenges to be awarded funding were announced on 10 May. See the guidance for public sector bodies on submitting a challenge.

3. Discussion

3.1 The technology innovation ecosystem

Government is a complex set of large organisations and the research shows that technology innovation is being explored in many different areas of the public sector. While government may not be considered an early adopter of new technology, the evidence shows that there is a large amount of experimentation with emerging technologies.

The research data collected was mapped into a set of emerging technologies that departments were using in prototypes, proof of concepts, pilots, trials and testbeds. Innovation networks, hubs and relationships with industry, academia, regulators and the tech sector were overlaid.

The technology innovation map shows what is being explored, where and by whom across central and local government.

The map is not meant to be a government organisation chart and does not cover all government bodies. However, it does indicate the central and local government departments and teams, where known, that are using new technologies and datasets to innovate.

The interconnections between government departments and with industry, business, academia, regulators and other international governments show the complexity of the government innovation ecosystem.

3.2 Emerging technologies used to innovate

The map shows which emerging technologies and data driven concepts departments, local and combined authorities are already exploring or which they intend to consider the potential of. These are grouped and listed below.

Table 2: The main emerging technologies and concepts that central and local government are exploring

Category	Emerging technology or concept	Description	Using or interested organisations
Automation	Artificial intelligence, machine learning (ML)	A set of rules that allows systems to learn directly from examples, data and experience.	Most central departments, local authorities, civil authorities
Automation	Deep learning	ML method that combines details for more abstract higher level features of the data using mathematical functions.	Cabinet Office (CO), GDS, Home Office, Ministry of Defence (MoD)
Automation	Robotic process automation (RPA)	Software that processes transactions, manipulates data from existing systems, triggers responses and communicates with other digital systems.	Large transactional departments, HM Revenue and Customs (HMRC), Department for Work and Pensions (DWP), Department of Health and Social Care (DHSC), MoD, Department for Education, Ofgem and local authorities
Identification and identity	Distributed ledger technology (DLT)	A decentralised (trust), resilient, immutable transaction ledger (database) that is cryptographically secure.	Defra, Food Standards Agency (FSA), Land Registry, HMRC
Identification and identity	Biometrics	Secure identity verification by automatic face, voice, fingerprint or iris recognition
Human computer interaction (HCI)	Voice assistants	Software that listens to voice input and acts on requests (links with AI for intelligent voice-based assistants).	Local authorities, DWP, Driver and Vehicle Licensing Agency (DVLA), Met Office
HCI	Augmented reality (AR)	Software that dynamically overlays contextual data and information on to real world views.	MoD, Met Office, Ordnance Survey (OS)
HCI	Virtual reality (VR)	Software that creates or replicates environments virtually.	DHSC, MoD, Department for Transport (DfT)
Internet of things (IoT)	Smart cities	ICT technologies, sensors and datasets combined securely to improve efficiency and manage cities’ services and assets.	DfT, local authorities, civil authorities, DCMS
IoT	Smart agriculture	Processes that use big data and sensors to improve agricultural yield and protect crops.	BEIS (agritech), Department for International Development (DfID), Defra
IoT	Smart homes	Reducing energy and water consumption with smart meters.	BEIS
IoT	Wearables	Sensors worn on the body to dynamically capture health, location, image and other types of data.	Home Office, DHSC, Ministry of Justice (MoJ), Environment Agency, local authorities
Mobility	Drones and unmanned vehicles	Aircraft and other vehicles controlled autonomously or remotely	MoD, Home Office (police forces), Defra
Mobility	Driverless vehicles	Robotic vehicle designed to travel without human operator	DfT and the Centre for Connected and Autonomous Vehicles, local authorities, civil authorities

3.3 Why departments are exploring emerging technologies

The research showed that departments and local authorities are exploring new and emerging technologies and opening up new datasets as part of their digital transformation programmes, efficiency programmes and manifesto commitments.

These programmes aim to improve the delivery and efficiency of public services and meet citizens’ rising expectations.

Further pressure on departments has come from EU Exit programmes and new regulatory requirements such as the General Data Protection Regulation.

Some departments have already recognised the value of testing new technologies by setting up dedicated innovation labs or teams, for example DWP, DfT, Defra, Met Office and UK Export Finance (by summer 2018).

Data labs and hubs have also been established, for example HMRC and Office for National Statistics (ONS), often with in-house analytics platforms, for example the MoJ, Met Office, NHS Digital and Home Office (HO). Data analysts and data scientists use domain specific data to gain new insights and, in the case of machine learning algorithms, new predictions.

Even without organisational structures like labs or hubs, development teams and enterprising individuals often spot ways to improve processes or test a new technology on their own.

GDS is using emerging technologies to improve GOV.UK

Machine learning:

• creating an increasingly accurate taxonomy of GOV.UK content by using supervised machine learning algorithms to correctly identify and tag more than 340,000 pieces of content helping end users find the right content, faster
• classifying user comments posted on GOV.UK webpages, using natural language processing saving manual processing time
• using machine learning algorithms to predict GOV.UK web page peak traffic demand on an hourly, weekly and monthly basis helping internal operations teams maintain service uptime

GDS is also exploring how voice assistant technology (such as Amazon Alexa, Google Assistant or Siri) can use GOV.UK content to provide speakable answers to user questions across the site. GDS is doing this to help make content as convenient to use and accessible as possible.

Centres of excellence (CoE) are teams or facilities that provide leadership, best practice and support for a specific technology or business concept. CO are leading a CoE for robotic process automation (RPA) with strong links to HMRC’s Auto Delivery Centre and DWP’s Intelligent Automation Garage which are dedicated facilities deploying RPA. HMRC is particularly experienced with many internal services and processes using the technology. Several departments are reviewing RPA for increasing productivity by automating administrative tasks.

Defra has an Earth Observation CoE that uses new technologies and satellite data to understand changes to the environment. GDS could consider new CoEs in line with its technical assurance and service delivery functions and data science capabilities. Those emerging technologies that gain more widespread use in government services will require best practice guidance. Examples might include CoEs for AI (partnering with the Office for AI), biometric technologies (partnering with HO) and distributed ledger technology. The Geospatial Commission is also considering a CoE for accelerating geospatial data use in the private, public and academic sectors.

Defence and health are major areas of research and development, backed by significant innovation funding. DHSC and MoD have specific innovation labs, hubs and programmes to understand how new technologies can be used across services. These departments also work directly with industry, regulatory and academic partners for innovation, for example DHSC and NHS England’s Academic Health Science Network Innovation exchanges, the Biomedical Catalyst, the Medicines and Healthcare products Regulatory Agency’s Innovation Office and the MoD’s Ploughshare, the Defence and Security Technology Laboratory and the JHub Innovation Centre.

Regulators are ensuring their approaches support innovation. A number of regulators have published innovation plans, for example the Health and Safety Executive, FSA, the Pensions Regulator, Ofsted and Homes and Communities Agency. The Financial Conduct Authority (FCA) and Office for Gas and Electricity Markets (Ofgem) directly enable and support technology innovation from tech firms in the market with regulatory sandboxes. These allow external businesses and startups to test new products and services in the market with real customers.

The FCA’s sandbox (as part of its Innovate programme) is considered world leading. It stimulates the financial services sector, attracts overseas businesses and fosters a public-private ecosystem. The regulator can supervise how new technologies (like AI) intend to be used in services and adapt regulation accordingly.

Innovate UK set up the Catapult programme that consists of 10 centres (catapults) to promote innovation between industry and academia with specific sector focus such as medicines, satellites and transport. The Digital catapult covers immersive technology (with an AR/VR lab), AI (with the Machine Intelligence Garage programme) and future networks (5G, LPWAN and IoT).

Barriers to exploring technology innovation were also expressed - why departments aren’t exploring emerging technologies. Bureaucracy, risk aversion, budget constraints, silo-thinking and innovation being a low priority given other demands were some of the reasons stated.

The impact of AI in government services

The AI Sector Deal was published by BEIS and DCMS in April 2018. This shows real commitment by government and AI has received a large amount of attention in government in the last 18 months. This is due to the breadth and depth of its potential impact on society and the economy as a disruptive technology. The rapid advances made in AI technology in the private sector and academic and consultancy research reports have fuelled this interest.

The AI review: Growing the Artificial Intelligence Industry in the UK published in October 2017 recommends that GDS develop a programme of actions to prepare the public sector and spread best practice for applying AI to improve operations and services for citizens.

And government should ensure that challenges addressed by the Industrial Strategy Challenge Fund and Small Business Research Initiative (SBRI) are designed to attract and support applications of AI using public sector data.

The AI Sector Deal also indicates that the GovTech programme could be used to support and deliver AI-based solutions to public sector challenges.

The GovTech programme (which uses the SBRI process) has received several AI-focused challenges from departments and local authorities for its first round of funded competitions in spring and summer 2018.

The research data shows that departments have started to explore AI use cases, but are doing so in isolation. There is an opportunity to connect departments all working on similar use cases, for example better fraud detection in transactions, predicting future outcomes and risk modelling, image recognition and so on (see table 5).

Despite AI’s enormous potential there are many risks and pitfalls with how government might use AI and associated datasets - for example:

• flawed underlying data used in training algorithms
• accountability and transparency of decision making algorithms
• liability of vendors selling AI algorithms used in government services
• fairness, bias and anonymity of personal data

The Office for AI, the Centre for Data Ethics and Innovation, and the Alan Turing Institute, along with tools like the Data Ethics Framework will help mitigate these risks.

3.4 How technology innovation is delivered

From a single developer researching a proof of concept to large multi-disciplinary teams working on testbed programmes, technology innovation is delivered in a variety of ways. Here are some examples from the survey.

Table 3: Examples of emerging technology based on use cases and how they are delivered

Delivery method	Developed by	Departments and other organisations	Emerging technology or data innovation use case
Proof of concepts, prototypes, minimum viable products (MVPs)	Departments’ development teams, policy advisers, academic partners, industry partners	HO, MoJ, CO Open Innovation Team (OIT), HMRC, Defra, DHSC, DWP, GDS, DfE, DCMS	DLT (securing digital evidence), DLT (border controls), DLT (verifying food provenance), VR (training tool for MRI scans), policy data visualisation (Churchill tool), streamlining statistical report process (RAP tool)
Proof of concepts, prototypes, minimum viable products (MVPs)	Innovation labs, data labs, DSA	DfT, MoJ, DfE Several, OS	ML (parliamentary questions analysis tools), ML (several examples from DSA cohort trainees), AR (overlaying geospatial landmarks)
MVPs / pilot trials	Industry (DeepMind) Industry (ASI) Industry (Atchai)	NHS Trusts, HO, DIT, DWP, DVLA, LAs	ML (better medical diagnosis eg cancer detection), ML (online video propaganda identification), HCI, ML (import law chatbot), voice assistant access to department specific services
Testbed programmes	Departments. with industry and academic partners	NHS England, Office for Life Sciences, DCMS	ML, IoT and wearables (multiple use cases), VR/AR, drones, IoT (5G Network, wireless rollout)
Sandboxes	Regulators with industry and startup partners	FCA, Ofgem	DLT, ML (several from Fintech and Regtech sector), IoT, DLT (energy storage and trading, smarthomes)
Government backed accelerators, catalysts, spin-outs	Industry, SME and startup partners	OS Geovation, Scotland - CivTech, CO - BIT, MoD - Ploughshare, CO - GDS GovTech Catalyst	Several including ML (multiple use cases)
Research and development funding or competitions	Industry, SME and academic partners	BEIS, UK Research and Innovation, Innovate UK	DLT, ML, IoT (several including energy, healthcare, policy-making), several including ML (multiple use cases)

3.5 Innovation networks

There are many formal and informal innovation networks both within government and linked to academia, industry and the tech sector. Here is a sample of some of the internal networks.

Table 4: A small sample of the many innovation related networks in government

Network or team	Host	Description
Heads of horizon scanning	Government Office for Science (GO-Science)	Large cross government group convening horizon scanning departments leads. Includes GO-Science, Public Services Innovation team, links with Cabinet Secretary Advisory Group on emerging technologies.
Communities of interest	Defra, DFID, OIT, HMRC	Community of interest for AI, community of interest on blockchain, informal groups that meet quarterly to discuss these technologies.
Transformation Peer Group	GDS, Infrastructure and Projects Authority	Cross government leadership group to discuss business transformation.
Future Policy Network	CO Economic and Domestic Secretariat	Wide cross government network of teams that focus on new approaches to delivery and policy making.
Policy Lab	CO	Group that supports and drives cross government policy innovation.
Digital Policy Leadership Group	DCMS	Formal group convening senior leaders on government digital projects and industry and market related digital themes.
DDaT Leaders Network	Various	Formal cross government network for convening Data and Technology and Digital Leaders to drive digital agenda, ensure technology delivery and ensure proper management and use of data.
Open Innovation Team (OIT)	CO	Links to academia and universities. Digital government partnership programme within OIT seconds PhD ‘Tech Fellows’ for 6 months.
Data Science Campus	ONS	Wide data analyst and data scientist network across government and academic institutions.
Knowledge Transfer Network	Innovate UK	Innovate UK’s network partner. Links to businesses, funders, universities and investors (used by GovTech programme).
Chief Scientific Advisers Network	GO-Science	Advise the Government Chief Scientific Adviser on all aspects of policy on science and technology.

3.6 The importance of data

Data has traditionally been a byproduct of operating services. Government recognised early on that data is a strategic asset. As a result, the UK leads the Open Data Barometer ranking for its open data principles and progressive stance to publish new, high quality datasets and registers. This enables data and technology innovation and fuels the data-driven economy.

Data is fundamental to emerging technologies. The new technologies in table 2 rely on data, either consuming it, generating it or both. Data is driving the rapid expansion of AI technology with predictive analytics, pattern detection, computer vision and natural language processing and departments have started exploring these techniques.

Massive data sets are often generated and at a faster pace than previously (for example from satellite imagery, IoT sensors and smart cities). This can cause issues around storing, processing and identifying the most relevant aspects within the data. Data privacy and data security continue to be of utmost importance when designing government services, including those that use new technologies, for example biometric authentication.

Data access, interoperability and sharing across government departments also needs to keep pace for new services or improving processes. The combination of emerging technologies and cross-department datasets being used together is often the enabler for a new innovation or insight.

Geospatial data is a highly valuable asset

Geospatial data has the potential to transform public policy - it is vital infrastructure. With the help of emerging technology, the pace of data innovation can be accelerated and improve how this data is accessed and visualised.

Machine learning makes data innovation at scale possible: detecting temporal, spatial and spectral changes at local, national and global levels within hours rather than days. For example, research is currently being carried out to determine whether coastal erosion patterns can be predicted using machine learning to enhance earth observation data received from satellites.

While some organisations have made progress by migrating their data to cloud infrastructure, the data tends to still be siloed with limited sharing. The Geospatial Commission strategy aims to improve access to and interoperability of this data for the public and private sectors, using federated and virtualisation technology.

This will unlock significant value and stimulate the creation of new innovative geospatial-based services.

3.7 Managing innovation risk

Experimenting with new technologies in government services and underlying processes is often seen as a risk.

The technologies may be untested, have no technology standards in place or are unable to scale. The stated benefits are not always clear and there can be a perception of new technology being a panacea.

The scale of experimentation also carries different risks. Testing a new technology to improve a small back office process is different to deploying disruptive technologies like AI or DLT in frontline services. Examples collected during the research range from low risk, low impact to high risk, high impact.

These are a few of the risks departments face when choosing to invest in technology innovation. But risks can be measured and managed and should be balanced against the risk of doing nothing, maintaining the status quo.

Some departments (see section 3.3) are actively managing this risk of inaction by establishing small scale innovation or data labs. Proofs of concept and prototypes provide ways to quickly evaluate whether emerging technologies can provide new, improved or more efficient solutions. Or, in the case of data analysis in data labs, new or better insights. Costs of failed experiments are also minimised and experience is still gained.

However, even with labs and hubs the research showed instances where similar solutions have been prototyped or developed independently. Or that departments were exploring the same technologies, use cases and approaches on their own.

Table 5: A few examples where coordination could minimise duplication and increase reuse across central and local government.

Use case	Departments or local authorities	Description
ML prediction	DWP, the Pensions Regulator, Ofsted, local authorities (Durham and Kent), BEIS, DHSC and NHS	Predictive modelling on datasets using different machine learning algorithms.
ML classification	MoJ, DfT, DfE	Automatic correspondence categorisation - for example Parliamentary Questions and analysis using machine learning algorithms.
Voice assistant access	Aylesbury Vale district council, DWP, DVLA, Met Office, HMRC, Hampshire county council, Public Health England, GDS, Welsh county councils, Hackney council, Norfolk council	Amazon Alexa and Google Assistant integrations for accessing government and local authority services.

Some visibility and coordination from the centre would be helpful. It would reduce the risks of departments reinventing the wheel and repeating past mistakes.

The GovTech programme also manages risk by using the SBRI process. Innovative technology solutions developed by GovTech firms tackling a public sector challenge follow a 2-phased competition process. The 2 phases (equivalent to ‘alpha’ and ‘private beta’) minimise the risk of tech firms developing incomplete solutions and ensures value for money. The programme also aims to mitigate the risk that developed solutions cannot be directly procured at the end of the process.

3.8 International organisations’ view of government innovation

International organisations like the United Nations (UN), World Economic Forum (WEF) and the Organisation for Economic Co-operation and Development (OECD) acknowledge the growing importance of how emerging technologies are increasingly intertwined with governance, policy and regulation.

They also recognise that the cycle of technology innovation moves faster than governments can grasp in order to create new policy. Research related to understanding and highlighting these issues is ongoing.

The UN is conducting a research study on Emerging Technologies for Digital Transformation in the Public Sector, to be completed in 2019. The research covers the emerging technologies already listed (see table 2) and will focus on the potential weaknesses, risks, opportunities and benefits.

The UK government is currently ranked the number 4 e-Government in the world by the UN (2018). The survey is conducted every 2 years. New for the 2018 survey, are specific questions to measure governments on emerging technologies, for example ‘do you utilise artificial intelligence, internet of things, blockchain, robotics, or other new and emerging technologies in delivering and managing online services?’

The OECD published a report in 2017 - Fostering innovation in the public sector. The report calls for governments to:

• invest in civil servants as the catalysts of innovation
• facilitate the free flow of information, data and knowledge across the public sector
• work together to solve problems and share risks
• use rules and process to support, not hinder, innovation

The OECD e-leader thematic group is carrying out a survey to discover how governments are using emerging technologies in the public sector. They are interested in how the technologies previously listed (see table 2) are used, particularly AI and blockchain. They have asked participating member governments to begin mapping these use cases. GDS has contributed to this survey with examples from the technology innovation map.

This work stems from the Observatory of Public Sector Innovation (OPSI) which recently published the report Embracing Innovation in Government - Global Trends 2018 where the UK (and GDS) are recognised for their innovation initiatives. OPSI is researching ways of measuring innovation-based outcomes and impacts to encourage the transition from incidental innovation to systematic innovation.

Related to this, think tanks like the Institute for Government measure international civil service effectiveness (InCISE) in governments worldwide. A range of metrics and attributes is used. One of the attributes it intends to start measuring moving forwards is Innovation.

3.9 Innovation capability

While the evidence is subjective, a subset of the attitudes from Nesta’s innovation competencies framework was observed during the research. Subject matter experts, DDaT leaders and technology and data specialists demonstrated some of the capabilities and motivation to drive innovation uptake in government. For example, during the set-up of the GovTech programme, individuals from stakeholder departments HMT, BEIS, DCMS and Innovate UK showed creativity and collaboration to iterate and refine the delivery model and process prior to launch.

Table 6: Nesta’s view on the skills, behaviours and attitudes required for driving innovation in the public sector

Innovation competencies	Definition	Skills and attitudes
Innovation	Experimenting and problem solving in the public sector.	Willing to take risks, action-oriented, resilient, outcomes and results focused, empathetic, creative.
Leading change	Mobilising resources and legitimacy to make change happen.	Political and bureaucratic awareness, financing innovation, advocacy, intrapreneurship.
Working together	Collaborating, engaging and creating shared ownership of new solutions.	Stakeholder engagement, building bridges, brokering, mediating, flexibility.
Accelerated learning	Exploring and iterating new ideas to inform and validate solutions.	Tech literacy, data literacy and evidence, prototyping and iterating, systems and design thinking.

There are a few technology and data innovation related training programmes in government.

The Data Science Accelerator is a cross-government training programme that originated as a partnership between GDS, ONS and Government Office for Science to build the government’s data science capability. The programme trains data analysts and data scientists to develop their skills in techniques such as machine learning, natural language processing and geospatial data analysis. The programme has now expanded to 6 regional hubs across the country, each hub hosted by a different department.

GDS and the DDaT profession have recently launched the Emerging Technologies Development Programme (ETDP) pilot. The 9-week programme trains developers to become subject matter experts in the latest technologies like AI. The training is provided by private sector companies and academic schools that are leaders in their field in applying the emerging technology. This builds capability within GDS and reinforces other GDS functions. After completing the course, graduates are seconded to departments to help advise on projects using the technologies.

4. Conclusions

Emerging technologies have started being tested in both central and local government to help support transformation, efficiency and EU exit programmes. They also support new regulatory requirements and the 2017 manifesto commitments.

Departments are exploring new technologies using prototypes, proofs of concepts and pilots to create better public services, improve internal processes and see what’s possible. Some departments have set up dedicated innovation and data labs to do this on a small scale. Centres of excellence are established to provide guidance and best practice for a specific area or technology focus.

Some findings are not widely communicated and there is opportunity to do more to share lessons learned. If not, there is risk of duplication, hindering reuse and of missing chances to jointly collaborate on common problems. In relation to emerging technologies it will be important for GDS to document best practice and update the Digital Service Standard and Technology Code of Practice as they are adopted across government.

Disruptive technologies such as AI and distributed ledger technology are receiving specific focus. They have potentially wide ranging process, policy and regulatory impacts on automation, identity, decision making, privacy, security and trust.

Several departments are investigating these 2 technologies but are often doing so independently and in a piecemeal fashion. The Office for AI and the AI and blockchain communities of interest groups are starting to address this, but GDS should do more to support and resource coordination and facilitate efforts from the centre to join up.

Technology innovation is also fostered between government, academia and the private sector. This includes technology-based research and development grants, funded competitions, joint academic programmes (for example the Engineering and Physical Sciences Research Council, Strategic Collaboration Funding, GovTech, Open Innovation Team Digital Government Partnership) and collaboration with government-backed accelerators and hubs (for example Geovation, CivTech). The GDS-led GovTech Catalyst team has built strong links with these accelerators and Innovate UK, who launch GovTech competitions to the market.

Understanding emerging technologies and how they are best used in government services is part of GDS’s capability. It is vital to embed and keep technical expertise in GDS teams so they can support departments using these technologies. The Data Science Accelerator programme and newly launched GDS and DDaT Emerging Technologies Development programme will ensure that capability is maintained and grown.

Combined with cross-government innovation networks, industry and tech sector links (for example FinTech, TechNation, Nesta, TechUK) the government innovation ecosystem is complex and diverse.

Other countries have national innovation strategies, and international bodies and foundations like the OECD, UN and WEF emphasise the increasing importance of innovation and disruptive technologies to governments.

The government’s Industrial Strategy challenges government to focus more on investment in research and development and innovation. Investment in technology innovation will be one important way that public bodies respond to this challenge. It’s encouraging to see recommendations on individual disruptive emerging technologies like AI, for example in the AI Sector Deal launched by BEIS and DCMS. The new GDS Innovation team has a key role to play in supporting BEIS and DCMS and government’s overall ambition to be a world leader on AI by leading on driving adoption of AI and other emerging technologies in government, to transform public services and improve public sector productivity.

GDS is in a good position to do this as it already provides advice and support on technology policy, service design and assurance to departments. GDS Innovation leads the cross-government GovTech programme which asks tech companies to make innovative use of emerging technologies to solve public sector challenges.

With GDS’s commitment to transformation and innovation, its deep DDaT engagement and its technical assurance, delivery and data science capabilities - GDS Innovation should lead work with departments to better coordinate, share best practice and drive technology innovation in government.

5. Recommendations

This report was commissioned by GDS. Below is a set of recommendations on next steps for GDS Innovation to lead and better coordinate technology innovation and the adoption of AI and emerging technologies in government - in support of BEIS and DCMS’s significant and much wider role across the economy in leading digital, technology, innovation and the industrial strategy.

The recommendations align to the new GDS Innovation team’s mission to support and enable the public sector to make innovative use of emerging technologies to transform services and increase productivity.

Table 7: Recommendations for GDS to better coordinate, share best practice and drive technology innovation in government

Theme	Area	Recommendation
Vision, leadership, strategy	Strategy	1. Develop and publish a cross government technology innovation strategy for public services by Q4 2018-19. Align with existing Transformation, Digital, Industrial and Cybersecurity strategies and 2019 Spending Review, working closely with industry and other government players (including BEIS, DCMS and HMT).
Vision, leadership, strategy	Innovation management	2. Build on GDS mapping work to date. Migrate the map data from the innovation ecosystem map into a queryable data and visualisation tool. Automate collection of future data points (for example using emerging technology - ML or RPA) Consider a formal emerging technology survey across government on current and planned use of emerging technologies. Use the map and survey data to identify and aggregate common emerging technology use cases, that are being independently addressed by departments.
Culture, community, capability	Training	3. Post-pilot, scale the GDS and DDaT AI and Emerging Technologies Development program for subject matter experts. Offer to all departments. Consider also running 1-day orientation learning courses of emerging technologies for senior non-technical government officials. Consider DDaT job family roles to include technologists specialising in innovation and emerging technology.
Culture, community, capability	Standards and assurance	4. Codify best practice use of emerging technologies. Update the Technology Code of Practice, Service Manual and Digital Service Standard to include guidance. Test with leading departments that are experimenting with the technologies.
Culture, community, capability	Centres of excellence (CoE)	5. GDS should consider taking more of a leadership and co-ordinating role on CoEs. Set up a network of CoEs to share best practice. Evaluate need with DDaT leaders networks, CoIs, the Office for AI and HO for new CoEs. Explore creation of new CoEs for AI and data science in GDS Innovation. Partner with HO for biometrics. For DLT evaluate from departmental survey results. Consider GDS sharing or hosting ML and DLT specific sandboxes as part of CoEs.
Culture, community, capability	Engagement and comms	6. Building on Sprint 18 innovation theme, organise and run an annual cross-government innovation conference (similar to Transforming Together). Communicate GDS innovation objectives and results on GOV.UK and GDS and use internal government comms channels to convene innovation champions and supporters
Projects, prototypes, products	Portfolio and exemplars	7. Based on common use cases from the technology innovation map data (see 2), build an innovation portfolio of emerging technology examples jointly with relevant departments, prototype joint solutions, work on common blockers and evaluate successes and failures.
Projects, prototypes, products	GovTech	8. Review first round of GovTech competitions in Q4 2018, evaluate ways to scale (more challenges and competitions) and align to BEIS business case. Work with BEIS and DCMS to deepen collaboration with private sector to encourage GovTech uptake and help build the GovTech sector.

Appendix: GovTech definitions and ecosystem

Table 8: Definitions of GovTech

Source	GovTech definition	Reference
pwc (consulting)	Driven by entrepreneurs using modern technology to disrupt the norm and deliver innovative products and services that users really want	Gov.Tech: the power to transform public services in the UK
Public.io (venture capital investment)	New technologies applied to public services and specifically designed for government purposes	State of the UK GovTech Market
GovTech Research (consulting)	GovTech solutions are innovative services which address a pain or need of government, citizen or business	GovTech: The sector set to revolutionise public services
pa (consulting)	Govtech provides solutions which offer new or alternative ways to deliver better public services and help government work in a more efficient and effective way	Internal CO: Govtech Strategic Market Review
CivTech (Scottish Government)	Civtech brings together public sector expertise and private sector creativity to solve real problems, develop new products, and deliver better, faster and easier services for everyone.	CivTech website