This blog is about things related to digital, technology and data happening in the Department for Transport (DfT) and with its partners. News and updates from the Department for Transport digital team.
The Transport Data Initiative (TDI) was founded in 2016. Set up by local authorities, its goal is to help its members improve the way they collect, store, and use data to make transport services better while reducing the costs of delivery.
The TDI is a partnership led by Buckinghamshire County Council and funded by Innovate UK. I sit on its steering group as the representative for the Department for Transport. The TDI runs regular events across the country to discuss data in transport. I was honoured to be asked to compère the 9th meeting in Sheffield recently.
An engaging event in Sheffield
The main themes of this event were:
updating the local authority community on the work on data that we are doing at DfT and, more widely, within the UK government network
highlighting the opportunities that data can unlock in transport
engaging with local authorities to discuss how the department can work with them to make these opportunities a reality
It was great to see the event full to capacity, with over 60 attendees representing councils across the country. Around 150 people joined online throughout the day via a live video stream and engaged with the event with questions and comments on social media. The conversations were engaging and challenging, with participants offering a lot of food for thought from a variety of perspectives.
An overview of digital and data in transport
The opening session highlighted how data and digital technologies had delivered customer benefits and operational efficiencies in other sectors. The Department for Transport is keen to see a similar step change in transport. This is reflected in the emphasis on data in the Future of mobility: urban strategy part of the government’s Industrial Strategy.
A summary was then provided of the work already being done by DfT and others – for example, the Joint Rail Data Action Plan, the Bus Open Data Programme, the Local Data Action Plan and forthcoming work such as the Transport Data Strategy and the Future of Mobility Regulatory Review.
TDI9 Sheffield - 2 DFT Data Overview MATT COLEMAN - YouTube
Local transport is on the brink of a revolution
Graham Hanson, Head of Smarter Traffic Management at DfT, outlined the progress being made with the Local Data Action Plan. Specific actions include:
the launch of a competition to help local authorities open up their data
an agile discovery to streamline and digitise traffic regulation orders, the legal instrument used by local authorities to regulate the use of certain roads
a feasibility study into establishing a metadata catalogue for transport data
local authority mobility platform research, exploring how urban traffic control management systems can be modernised to provide multi-modal open transport data and encourage third party innovation
TDI9 Sheffield - 3 Local Data Action Plan GRAHAM HANSON - YouTube
Darren Capes, ITS policy lead at DfT, and John Cooper, senior ITS engineer at DfT, provided more detail on the local authority mobility platform research. This is a project being undertaken by the Connected Places Catapult, which involves developing use cases such as:
Challenges and opportunities are becoming increasingly clear
There was an ‘ask DfT’ session, where department officials responded directly to queries from the audience. The variety of topics and angles covered was intriguing:
how best to get third parties to share their data
the importance of revenue funding as well as capital for innovative initiatives
the need for commercial and economic analysis of new business models to encourage data sharing
the potential role of crowdsourcing data
new mobility modes and micromobility
the challenges with procurement in innovation
Many of these topics will be examined as part of the Future of Mobility Regulatory Review and the Transport Data Strategy.
How to get involved
There are several ways to get involved.
The first is to contribute to the final session of the event, which has an online counterpart. In the session, Abbas Lokat, a senior consultant with GeoPlace, hosted a workshop on the traffic regulation order discovery project, which seeks to understand:
the costs associated with the creation of traffic regulation orders
to what extent the features associated with these have been digitised
A second way to contribute is to attend the next Transport Data Initiative event, the 10th. It will be held on 19 June in Bristol, and will focus on procurement, skills, and new business models – please check the TDI website for registration details.
Bernadette Kelly, Permanent Secretary, Department for Transport
Earlier this month, colleagues from across the Department for Transport and its agencies gathered for an inaugural International Women’s Day themed event. The aim was to:
showcase the power of diversity through inspiring role models and their success stories
really help understand what it means to work in Digital Data and Technology (DDaT) roles in DfT
The afternoon kicked off with a rousing introduction from DfT’s Permanent Secretary, Bernadette Kelly, herself a fantastic role model and one of the few female permanent secretaries in government. She expressed concerns that, although the Civil Service has come a long way, there is much work still to do.
Although there are increasing numbers of women employed across government, this drops significantly when you get to senior levels. Bernadette is particularly keen to address the gender pay gap within DfT and emphasised the importance of ‘never taking your foot off the throttle’ when striving for diversity in the workplace.
Attendees were privileged to hear next from three women who have carved out hugely successful careers in the digital sector: Rachel Murphy and Karen Cleale from digital consultancy Difrent and Emma Stace, Chief Digital Officer at the Department for Education.
Their talks were honest, moving, impactful and stuffed full of helpful tips based on their own experiences for women making their way in the digital world. All shared intensely personal experiences that have helped shape them and their careers. Their stories included:
finding the strength and confidence to stand up to those at home or at work who are not supportive
bravely striking out as a single mother
finding inspiration from a stepdaughter who fought back from a life-threatening illness.
Emma talked about how she has experienced unhelpful comments as a woman, including being told how to dress and being accused of being too emotional in the workplace. She asked the audience to chat to each other for a few minutes about sexism that they have experienced at work.
It was clear from the discussion that followed that sometimes these comments come from female as well as male colleagues. Sadly, women are not always supportive of other women in the workplace.
A fireside chat was next on the agenda featuring inspirational women from inside and outside DfT: Barbara Keating (digital architect, DfT), Renate Samson (senior policy advisor, Open Data Institute), Yalena Coleman (Head of Accounts, Transport Systems Catapult), Jan Ford (Head of IT project delivery, HS2), Sarah Winmill (Chief Information Officer, British Transport Police), Sunitha Chacko (Head of Technical Architecture, Government Digital Service). They shared their own fascinating career stories and experiences and the audience had the opportunity to ask questions.
Some key messages that we can all take away:
Don’t be afraid to apply for jobs in digital. You don’t need to be a techie, you don’t need to have all the skills. Just go for it.
Be authentic and honest. As a (male) colleague once said to Emma Stace, ‘You are enough.’ Often people have more respect for leaders who display these qualities.
A recurring theme of the afternoon was imposter syndrome, in which we doubt our accomplishments and fear being exposed as a fraud and this translates into some of the language we use at work. We should stop saying ‘sorry’, ‘just’, ‘does that make sense’ and have confidence in our knowledge and abilities and what we are saying.
Rachel Murphy emphasised how we should also be supported by men. She cited her father as being instrumental in encouraging her that she could do anything and her business partner in providing essential support and advice. Jonathan Neffgen, CIO at DfT, talked about how he mentors women (and only women) via Linkedin to help them to develop their digital careers.
International Women’s Day might only be one day each year but it is essential that all of us keep working hard every single day to ensure that women have equal opportunities, are properly represented and fairly treated within the workplace.
I work on the Platform Team of the MOT testing service. We build the cloud infrastructure that the application runs on and the automated release pipelines that get code from a developer’s laptop to the live service via lots of rounds of testing. The service is there to support our users (MOT testers) in their work. My role is to make sure that the service runs smoothly and to support our developers in making the service better.
My background is in science and I came to DVSA 2 years ago as part of a graduate scheme having zero experience of agile software delivery. The DVSA put a lot of effort into training me up and I can’t believe how far I’ve come in that time. What I love most about my job is that no 2 days are alike. On the platform team we use a multitude of different technologies on a daily basis so there is always something new to learn. I also get to flex my problem solving muscles because, when things go wrong in unexpected ways, part of my job is to work out why.
Getting more women, and generally more people from a wider range of backgrounds, into digital roles is essential because diversity helps everyone make better things. I know first-hand what it is like to grow up in a house with no internet access and therefore how important it is. As government employees, we take seriously our responsibility to build services for everyone and not leave anyone behind. One of my computing heroes, Grace Hopper, spoke of the danger of the phrase “We’ve always done it this way” and that is just so true. If you want to be an innovator, if you want to agitate and to build incredible things, you need to surround yourself with people who challenge your assumptions every day. And that’s why diversity is so, so important.
Lindsay De Bank, Head of Project Delivery, DfT
I joined DfT in January 2018 from the private sector. My first task was to build a new team to meet the department’s increasing demand for digital and technology projects. The team includes delivery managers, project managers, business analysts and a software development team.
My role is to set the standards, frameworks and clear expectations that allow multi-disciplinary teams to operate successfully and deliver at pace.
We offer lots of opportunities for apprentices at the start of their digital careers, so creating a safe environment to learn, and the support to do this whilst delivering, is critical. It’s a challenge I didn’t experience in the private sector. No 2 days are the same and I love that variety and uncertainty. My day could involve:
• urgent problem solving to remove blockers that are holding up a project
• supporting project teams in managing stakeholders
• negotiating contracts
• coaching and mentoring the team
• facilitating cross project/DfT workshops
• recruiting new staff
• attending project boards and investment committees
I do what I can to provide a clear pathway for teams to deliver new user-centred services successfully, quickly and with least resistance, learning from one another and beyond.
DfT regularly wins awards for diversity as an employer, but there is still a way to go. I’m currently the sole female on the DfT Digital Service Leadership Team of 7 people and 1 in 8 members of my team are currently women. I’m a firm believer that people should be recruited and promoted based on their ability, rather than their gender or any other characteristic. That said, I’ve met a lot of great women in the digital arena throughout my career, and I’d love to see more in my teams and across the Digital Service at DfT. Whilst it’s great for applicants, when we’re looking for a balanced panel at recruitment or a range of different perspectives on projects, I suddenly find myself in great demand!
Anna Davies, Lead User Researcher, DVLA
My work within customer insight and user research has definitely been the most rewarding during my 15 year career at DVLA, and was not an area I had known much about previously.
Being the bridge between our customers and the service designers, who are tasked with providing services which are simple to use, is an interesting one. It’s taught me never to make assumptions! I provide guidance and support to my team of 7 user researchers. Our aim is to gather purposeful insight which can be used to improve current digital services and develop new ones. It’s a role I thoroughly enjoy, engaging with a variety of stakeholders both internally and externally. The culture at DVLA is changing and I am lucky to be given full autonomy to lead my team as I wish.
I am keen to develop my skills in both leadership and user research using the opportunities provided to me at DVLA, furthering my career in the organisation. It’s a great place to work and I’m hoping that the open and inclusive culture will continue to grow, providing more opportunities to women in the future.
Liz-Ann Rodrigues, Digital Architect, DfT
I joined DfT from the private sector as a digital architect in January 2018. My role in a nutshell is to provide technical leadership across the department by being a strategic thinker, technical expert and also being empathetic in my approach.
To me, this particular role is an opportunity to make a difference to people’s everyday lives. Working at DfT has provided me with many such opportunities. Every day, I know that I am working for the good of others, whether I am improving capabilities within my department or designing new citizen-facing services.
There is a lot of research which proves that women bring different perspectives to the table. I feel welcome and confident in my work at DfT. I think the range of people from different backgrounds helps us to design better services and makes DfT a great place to work.
Vanessa Hughes, User Researcher, DfT
I joined DfT as a user researcher in the Digital Adoption & Innovation team 6 months ago. Before that, I spent 15 years working in DWP's Child Maintenance Group. My role is to interact with development teams and other project stakeholders to create a shared understanding of users and their needs. It takes an empathetic approach to understand the constraints of the team and to challenge them to become more user-centred and take on research findings.
I joined DfT for a new challenge in a new environment. In my previous role I experienced close up the impact government services can have on people's everyday lives, and the difficulties colleagues face in using the systems to deliver them. It’s fantastic to be part of a team bringing user-centred design to the digital platform, which increasingly shapes the way society interacts with government.
This being my first role in the digital sector, it felt both daunting and exciting. Moving from a relatively female dominated department, I wondered whether I would fit into the
male world of digital, data and technology. But, in my experience, there is no such dominance. I haven't arrived as part of a mission to assert the female gender into DDaT, I've joined a melting pot where people pool their skills and theories to produce something new. Digital is the tool; you are the welcomed facilitator, the communicator, the empathiser, the innovator. The more diverse the skill set the richer the soup, so I'd urge you to come and add your ingredients.
Jo Moorshead, Head of Knowledge and Information Management, DfT
Information is the lifeblood of what we do. In a paper world, managing information was relatively straightforward. The digital world brings both opportunities (the richness of the information that is available to us and increasingly sophisticated tools to help us discover and use this) and challenges (how we go about managing this huge volume of digital information and ensuring it survives for as long as we need it). But that is what excites me and motivates me to come to work each day. There is never a dull day in Digital Service!
I’m proud to be a woman working within Digital Service. Although a cross government profession in its own right, KIM is very closely aligned with the DDAT profession and there are strong overlaps between the two areas. In my experience, there are more women in the KIM community - the KIM team at DfT is entirely female right now. As these two professions work more closely together and learn from each other this opens the possibility for more women to enter digital roles.
As a mum of two daughters I’m keen to be a positive role model and have worked hard to balance my role as a parent with a fulfilling career. My message to them is to never be afraid to do what you want to do and try new things. You don’t have to be the best, the important thing is to give it a go.
Latest Earnings Networked Nationally Overnight (LENNON) is an application used by the rail industry. It provides data, such as ticket sales and franchise earnings, which helps them better understand how the rail network operates. The application is hosted externally, but DfT maintains an in-house version containing a subset of the data for our own analysis.
It’s a huge system at more than 100 terabytes. And it’s the most heavily used system by our Rail Technical and Data Management team.
Slow query times on an application like this can be a real nuisance for those relying on frequent access to the data. And this lack of speed means we can’t exploit the data in the system to its full potential.
Google Cloud Platform
So what could we do to speed the system up and allow us to make better use of it?
As part of a wider transformation of the digital technology in use at DfT, we undertook a discovery exercise with Google. This gave us a clear route to shut down our in-house data centres, move our services to Google Cloud Platform (GCP) and transform existing products onto GCP. LENNON was one of these transformation candidates.
After the Google discovery, we set up a Cloud & Data Centre Transformation project. Google, their partner CTS and DfT’s Digital Service, alongside the Rail Technical and Data Management team, were tasked with speeding up slow query times and improving the system’s capabilities. Not an easy task, but certainly possible.
And that’s not all we were hoping to achieve. Backups and maintenance on GCP should be ‘frictionless’. By comparison, the current application requires frequent manual intervention from the colleagues using it. Moving it to GCP will free up time and resources that could be better used elsewhere. The transformation should also increase the security of the application, something never to be overlooked.
This would be our first transformation of an existing application onto GCP, hopefully proving it as a reliable, cost effective and efficient platform for data processing among other things. So how has it gone so far?
LENNON leads the way
Although the project is still in progress, we have already seen some clear and valuable benefits. According to the Rail Technical and Data Management team, processing speeds have fallen dramatically. Where it used to take several hours to execute a query, it is now taking less than 20 seconds.
Not only this but the application can now run multiple queries simultaneously without affecting its performance. These improvements have been a huge help to our colleagues in Rail IT who can now consistently receive the correct data, quicker.
From the practical side of things, there is now also the ability to use DfT’s instance of LENNON for all of our queries. Previously, we had to rely on Rail Delivery Group’s version of LENNON to run some queries. Removing this dependency will save time and effort both for us and RDG.
For DfT’s Digital Service, this project has demonstrated some huge benefits of using GCP. I hope and expect that this experience will enable us to further improve our efficiency through the transformation of similar applications.
In the words of our Chief Architect, Mark Lyons:
"The transformation of LENNON within GCP is a really exciting first step towards moving and transforming our on-premise services into Google Cloud Platform within the next six months. This has been a fantastic example of collaboration between ourselves, Google, their partners CTS and our Rail colleagues, and it has given us a design pattern that can be re-used for other data processing and analysis requirements across the wider department"
Looking forward, we have already started to reuse this approach for other projects in their early stages. It’s exciting to think that data processing as powerful as this should be available for even more projects down the line. Here’s to the cloudy future!
"Is it the unit or the data that is developing?" I've been asked this, jokingly, a few times since I started as Head of the Developing Data Unit (DDU). I've always replied, also joking, "Both, of course!". Yet I realise there is some truth to my reply.
The DDU is a new central team within the Department for Transport (DfT). Our remit is broad, encompassing all things data within DfT. As such we are a developing unit. At the same time, data is a ‘developing’ concept by nature, especially in transport.
We are a cross-functional team, working in two broad areas:
Data strategy and culture: We engage across DfT to develop a data-driven culture, and to define and implement a data strategy. To do this, we work with colleagues in all parts of the organisation and the wider transport sector.
Operational support to data-related projects: We offer ‘data intervention’ across the spectrum of data issues: technical, ethical, commercial, governance.
Transport is evolving
This year marks the 100th anniversary since the then Ministry of Transport was set up. Transport has changed a lot over the last century.
More recently, multi-modal transport has become easier thanks to transport authorities releasing their data. Innovative apps like Citymapper use that data to help people to plan their journey, offering choices they might otherwise have been unaware of.
This type of innovation is driven by data. Previously, data was only used to produce analysis and statistics after the fact. Today, live data is increasingly used to transform the way people travel.
Data can also enable policy makers and operators to make better decisions. It can be used in transport simulations to more accurately predict the effect of a course of action. It can help the transport network be more responsive to disruptions, reducing the impact on travellers and businesses. It can enable the private sector to develop innovative journey planning services.
In this developing data context, the Department for Transport:
facilitates data sharing between local transport authorities and operators
makes the case for policy, legislation, or investment to unlock value to transport users and operators
Banned: data for data's sake
As a team that is working to give data a more central role, we want to have an impact on all the above areas. We will work to make sure that DfT is well placed in this data-driven arena. We will support both internal and external data work across the spectrum of transport providers and users.
But we don’t want to forget that data is a means to an end. People, who make travel decisions based on data, must be at the centre of our thinking. Ultimately, we want everyone to enjoy better transport services, and the public and private sector to provide them in a way that benefits everyone.
map and catalogue available datasets in transport, whoever holds it
help DfT build its data culture
release as much data as we can and help others do the same
develop publication standards
streamline and document data release processes
make sure that data is not an afterthought but equally not a hurdle or a box to be ticked
make sure that data ethics is embedded in our thinking
We’re supporting colleagues in several areas. For example we are facilitating the publication of bus open data, helping the Smart Traffic Team evaluate their Local Authority Data Innovation competition, reviewing options to collect roadworks data, discovering rail data.
The Developing Data Unit is not here to play buzzword bingo, but it is in our DNA to wear many hats, as both generalists and specialists. Specialists in data, but with a generalist, agile approach to the many ramifications of data workstreams: strategy, policy, openness, ethics, digital architecture, and physical infrastructure.
This is reflected in the skills and expertise of the team. We have data policy and strategy experts, technical specialists, data management leads. We work closely with our colleagues in the digital and data analytics teams and more widely with other analysts and project leads on data projects, and with partners in the transport sector.
The Web Foundation recently published an interesting article, noting that open data is still in beta, 10 years after it appeared on the agenda of governments and local authorities around the world. In my view, the underlying issue is that data processes often have not yet become embedded as business as usual for many large organisations. Data releases abound, and that is a positive thing, but important questions around data have largely gone unanswered. We need to address issues such as how to:
make data releases sustainable
make sure that data contributes effectively to operations and contracts
gain the constantly evolving skills required to deal with data
Obviously not all data can be open. But the lessons we learn from open data can teach us useful lessons on the road to better services, better policies, and better deals, in a way that benefits operators and the public alike.
This is what we intend to work on. We are defining our priorities for 2019. It's going to be an interesting and challenging year. Please get in touch if you want to discuss transport data with us!
We are recruiting for a Strategic Data Manager. Are you an experienced data professional? Would you like an opportunity to help shape the transport sector’s use of data assets? Do you want to work on projects that have the potential for greatest real-world impact? If so, we would like to hear from you! View the job advert.
Earlier this month, we ran the first ever DfT Group Digital, Data & Technology (DDaT) Profession Conference. Not only that, but it was an ‘unconference’. I’ve written this blog to explain why we did it, what happened on the day and how you can run your own unconference.
The Department for Transport has—or works closely with—22 agencies and public bodies. The DDaT profession is a young one across the DfT group. With all the professions in DfT, we look at 6 areas for development:
talent management & career pathways
curriculum & qualifications
standards and competence
When it comes to networks, we already have a regular meeting between the most senior leaders in digital from some of those organisations, but there’s no formal contact at other levels.
We decided to change that by organising an unconference. So a group of us from the Department for Transport, the Driver and Vehicle Standards Agency, Government Digital Service, High Speed 2, Highways England, the Maritime & Coastguard Agency and the Vehicle Certification Agency got together in Bristol.
An unconference is a great way to get people to network, without it feeling like networking. Participants pitch topics at the start of the day, vote on their favourites and run their own sessions. When done right, everyone should find everything they attend relevant. If they don’t, they simply walk out and go to another topic. This video of the day gives you a flavour of what it is like.
DDat UnConference video - YouTube
We had up to 6 topics running at any one time, with topics closing and opening throughout the day, so we covered a lot. Each topic got its own hashtag so we could live blog about them (more on that later). Here’s a selection of what we covered.
#betterdata – How can we better organise and use data efficiently across the business? #disruptive – What impact will disruptive technology have on various modes of transport? Are we gearing ourselves up as an organisation to meet the upcoming challenges? #diversity - How can we increase diversity in DDaT roles? #excelisubiquitous – How do we overcome siloed data stuck in excel spreadsheets? #innovation - How do you encourage innovation when governance can make it challenging? #networking - How can I contact people in similar roles in other DfT family organisations? How do we share our knowledge? #nextchapter – What can we do to transform transport with digital? #O365 - How are people using Office 365 in their organisation? #sharingiscaring - How can we make sure that DDaT roles outside central digital functions are well supported? #training - Would cross-organisation training be a good idea and, if yes, what can be done to achieve it?
I’d never even been to an unconference before. If I can organise one, so can you. We started with some basic principles as set out by others here, here and here.
How to get good topics
Although we didn't do so on this occasion, you could set a theme for the day, or pose a question. For instance, I would like to run another unconference with DDaT and policy professionals posing the question “How can we use digital in transport policy?”
The beauty of an unconference is that one person might interpret that as “How do we use digital to transform how we make policy?” and suggest as a topic “How do we use machine learning to get better insights about how people use the transport system?” Another person could think it means “How do we bring about a transport system that makes the most of digital?” and propose “How will the economics of data impact the transport sector?”
Get people thinking about what they want to discuss as early as possible. We did registration through Eventbrite and asked participants what they wanted to get out of the conference. Once we had a list of topics, we sent the full list back out to everyone before the event.
We also asked people to answer three questions before they arrived:
1. What’s a success (or failure) you have had that others could learn from?
2. What are you finding hardest to solve right now?
3. Who else are you hoping will be at the unconference (could be someone with a particular role, skill set or experience)?
Assume your participants don’t know how an unconference works and explain the principles several times before, at the start and during the event. At first, it might feel awkward, so you need to encourage people, but also give them feedback on how they can get more out of the unconference. Firstly, we had to keep pushing people to follow the law of two feet.
"If, during the course of the gathering, any person finds themselves in a situation where they are neither learning nor contributing, they must use their feet and go to some more productive place." from Transition Culture.
As well as ensuring that everyone is only staying in topics relevant to them, this principle also prevents zombie topics that refuse to die. Once no one is at the topic anymore, you can kick off a new one.
Secondly, keep an eye on group dynamics. At the start, it was clear that the people who pitched the topics felt they needed to lead the conversations. After the first break, we asked participants to set their chairs in circles and have everyone either sit or stand so that no one person had more control of the conversation.
How technology can help
Technology made everything a lot easier, although always have a supply of sticky notes to hand in case it fails. We used Slido to enable everyone to pitch topics simultaneously. This is a great way to ensure you hear introverts’ ideas too. We asked people for a question, a hashtag (for live blogging) and their name (so we could ask them to kick off the conversation). Everyone could then upvote what they were interested in. We also used different rooms on Slido to host our live blog, an up-to-date feed of what topics were happening in each room, and to document the commitments we all made at the end of the day.
We also used WhatsApp to stay in contact with volunteers in each room. When anything changed throughout the day (and given the dynamic nature of an unconference, that happens a lot) we could let them know immediately. It also gave us a way of letting everyone know when a new topic started, as the volunteers would announce them to the room they were in.
How to ensure that it’s not all talk
It's important to finish the day with a focus on actions. For our last session, we had everyone together in our largest room. All the topics from the day were spread out on the walls. We got people to write down what actions they wanted to take forward and put them up next to the relevant topic title. Finally, we got people to congregate around the action they care most about and discuss how they could work together on it. The law of two feet still applied, so people could still move around. We finished the day by all putting our shared actions on Slido.
My commitment was to run another DfT Group DDaT Unconference. In the meantime, a few people have already committed to setting up a social media network so we can broaden the conversation and convert our discussions into actions, which I'll share in a future blog. Watch this space!
DfT’s Rail Knowledge Management Team recently approached the Lab for help with a problem: new DfT employees needed to understand the complexities of the rail system. The rail team’s data suggested that site visits and tours were an effective way of building understanding of how the pieces of the railway work together, but these events happen infrequently. We wanted to prototype a solution that allowed all staff to access these opportunities at a time that suited them, and so ‘Virtual Raility’ was born: a virtual reality training tool for new staff.
Once the team had created the first version of the prototype, we began usability testing. Virtual reality was new to us all, and none of us were sure what a good user experience should look like or how to effectively test. So here’s what we learned:
Seeing what the user sees
Due to the availability of screens, we weren’t able to set up Chromecast or something similar that would allow us to cast what the user was seeing onto a bigger screen. To get around this, we ensured our discussion guide placed extra emphasis on asking users to ‘think aloud’ to describe what they were seeing, and asked users to complete tasks that would also allow us to understand where they were in the service. In future we’d make sure to arrange testing in a room with a screen and Chromecast to project their view.
Text or audio?
Users generally found the experience to be exciting, but struggled to take in the textual information in our first prototype. We learned that it’s very easy to overload the visual realm of VR, and subsequently added audio content to all our interaction points. This also made the app more usable for users with accessibility needs. We learned that people’s eyesight and hearing is highly variable, so next time we’d do a lot more testing, and ideally begin to implement settings for the user.
Making the journey clear
Users interacted with the app by focusing their gaze on hotspots that would expand and provide textual information on what they were seeing. Our first iteration made all the hotspots visible from the beginning, allowing the user to access whatever they wanted from the start. However, this lump of visual information overwhelmed the user and prevented us from applying a narrative to the content. Our next iteration guided the user, hotspot by hotspot, using a floating breadcrumb trail between points. Users much preferred this approach and it allowed us to provide a more structured learning experience which improved information retention.
Not all users can tolerate virtual reality for long
We read about users with less experience of using virtual reality having a lower tolerance for the technology – and our usability testing confirmed this. While some people were fine to go through the experience for 3-5 minutes, others stopped early and chose not to proceed. These users usually complained of feeling dizzy or giddy, or their eyes feeling a little tired. We learned a few lessons to reduce this affect:
keep the users’ vision at head height
place text at a consistent depth in the users vision
keep the experience short
Get the motion right
One of the hotspots inside our service transported the user from Victoria Station to Clapham Junction. Earlier iterations of the prototype didn’t give the user any notice that they would be transported, so many found that they looked at a hotspot then suddenly moved through space to another station. We started affectionately referring to it as our Harry Potter-esque Floo Network and users found it variously exciting, startling and disorienting. Our later iterations of this interaction gave the user some notice of what was about to happen, and transported them more smoothly to the next scene.
If you’d like to learn more about how we built Virtual Raility in just 4 weeks, drop us a line at firstname.lastname@example.org
In my previous Blog, Applying data science to policy, I talked about our first attempt at developing a data science software application to automate some painful parts of the policy consultation process. I alluded to challenges we faced and the fact that we could learn from these to deliver a better product and become a better data science team. This post gives a little more detail about our experiences. It’s basic stuff, but it’s easy to lose sight of that when you’re neck deep in code.
Keep it simple
One of the reasons the first iteration took so long is that we fell into the ‘shiny thing’ trap. As we developed, we encountered issues and we were often presented with 2 choices. We could do something quick and simple that resolved 80% of the issue, or we could invest the time and effort into a more complex solution that attempted to solve 95%+ of the issue. Often, we took the second path without trying the first.
One good example is around the removal of ‘junk text’ of no value to the analysis (for example signatures, virus scanning text). We spent a lot of time building a machine learning algorithm to solve this issue. It didn’t work as we’d hoped and when compared to a simpler pattern based system (regular expressions to the coders among you), the simpler system performed a lot better. The lesson? Simplicity often beats complexity. Start simple and iterate and beware the allure of shiny new things.
We also unwittingly adopted the ‘LGTM’ mentality. LGTM stands for ‘Looks Good To Me’ and is often used in software development circles to indicate a ‘light’ code review. Some of our code didn’t work as intended when we deployed it. This resulted in a lot of time identifying and resolving bugs and rewriting code, particularly for edge cases.
So, we’ve now adopted stricter code review processes and coding standards with a ‘little and often’ mentality. We supplement this with regular ‘show the thing’ sessions to review and learn from each other’s code and stricter testing principles to make sure that our new code doesn’t break the old code. There is still more to learn of course and it’s easy to think of this time as unnecessary. After all, reviewing someone else’s code isn’t fun. But it sure beats fixing someone else’s code. Or worse, fixing your own code from 6 months ago.
One of the things we got right was around deployment. Adopting a Cloud First Policy, we developed the application to be hosted in the cloud. This gave us a lot of freedom and presented new challenges. How could we quickly and easily deploy and manage our application which consisted of numerous databases, components and interconnected parts? Doing so manually would have taken us around 3 hours of downloading, deploying and configuring each time.
The solution was a technology called Docker. This was more work up front as we had to both learn and configure Docker for our application. But, once it was up and running, the benefits became clear:
deploying the application was as simple as a few lines of code which pretty much anyone could do
it made our application more shareable and reusable
it saved us literally days of time during the development, testing and deployment process
having learned it, we’ve applied it to other projects
Docker (https://www.docker.com/) is also free and open-source and widely used with a myriad of documentation and examples.
Always be data sciencin’
The last takeaway was around the role of the data scientist. At the time of building this, our digital department were undergoing a transformation to be more focused on innovation. They supported us with hardware and a cloud platform which was literally game-changing in terms of the possibilities it presented. But they didn’t have people in place yet to contribute, so we used a contractor and learned things ourselves.
It was fun to learn how to build web applications and we could apply these skills to our other projects. But this shouldn’t be at the expense of core data science skills including data processing, analysis and machine learning. One big lesson for me personally is that it’s very easy for your skillset to develop around projects. You should step back and take a more strategic view otherwise before you know it you’ve turned into a software developer.
So, for this year I’ve come up with a data science oriented development plan and I’m dedicating time in my diary to it. I’ll no doubt learn wider skills on the job too, and I’m fine with that as long they’re ‘as well as’ and not ‘instead of’ core data science skills.
Transport is a key part of most people’s day and so it inevitably ends up in the news. Much of this news appears in local papers, which make up the majority of the 700+ news outlets in the UK. For some policymakers, these news stories are a useful and informal input to the policy development process.
As human readers who understand how the English language works and what the Department for Transport (DfT) does, we can quickly identify if a piece of news is relevant to our policymakers. Yet, despite our suitability to the task, there is simply too much news generated each day to keep up with. This got the Data Science team thinking: could we teach a computer to do this for us?
The problem with keywords
A common approach to solving this is to come up with a list of keywords you’re interested in and search the news media each day for articles that contain these words. But as these news results with the word ‘car’ in the headline show, this approach quickly runs into problems:
“Pranksters put HALF a Mini car at top of the 'Nuneaton nipple' - but why?” Coventry Live
“Drivers are angry at queues of 'more than two hours' to get out of St David's car park in Cardiff” Wales Online
“Highly charged: complaints as electric car points block city pavements” The Guardian
What has been placed atop the ‘Nuneaton nipple’ is probably of little interest to colleagues at DfT, but the other 2 articles may be of interest. So to refine the results, we need to use keywords in a smarter way.
One way could be to write more complex rules to filter the news. For example, the Plug-in Vehicle Infrastructure team could receive articles that contain the words ‘car’ and ‘electric’ or ‘car’ and ‘charging’. But this doesn’t get to the root of the issue - there are a vast amount of niche topics that different teams are interested in, and manually programing the right filters would be a mammoth task. Plus, the programmers don’t have the subject matter expertise to come up with sensible filters.
Our solution is to outsource the subject expertise to the teams themselves. This gives them an easy way to train the computer to know which articles are relevant today, based on what was relevant in the past. And using machine learning algorithms to transfer the heavy processing to the computer.
Our current prototype addresses the core problem. We used 300 articles that we already knew were relevant to aviation policy to teach the model to identify new aviation articles correctly.
Computers aren’t equipped to deal with language, but they can process increasingly vast amounts of numbers. To convert text into something a computer can digest we convert sentences into a matrix of word counts, referred to as a ‘bag of words’. If we do this for a selection of the words featured in the car headline results, we can begin to spot patterns in the data. For example, the Guardian article headline features the words ‘car’ and ‘electric’ but not ‘park’ or ‘queues’.
Word counts table
If we have a selection of articles that we know are of interest to a particular team, we can use the word counts of these articles to train a model, such as a Naive Bayes classifier (PDF 326KB), to make predictions.
Without going into detail, the process involves calculating the probability that a word, for example ‘park’ will appear in documents we’ve tagged as belonging to a team. And using those probabilities, it can calculate the most likely category for new documents. We can see from the word counts table the probability that articles with the word ‘park’ belong to the Parking category is much higher than the probability for the word ‘electric’. In reality, we apply this process to thousands of words and hundreds of articles.
Once the model has identified relevant articles, we use natural language processing to select the most important keywords from the article and the most important sentence. We include this in a daily email and send it to the relevant team. In this way, we hope the recipient gets a summary of what the article is about without having to open the link (just in case the headline is not accurate).
In the long term, we want to make the tool self-sufficient, by allowing recipients to improve the predictions with each email by voting whether the selected article was relevant or not. Doing so will provide the model with more data which in turn will improve the accuracy of predictions.
Get in touch
We are also exploring other ways we can use this data - at present we have 40,000 articles that contain transport related keywords. If this project is of interest to you - or you’re interested in other text processing get in touch at Data.Science@dft.gov.uk
One of the great things about working at the Department for Transport (DfT) is that it’s a really outward-facing department. During my 3 years here, I’ve met people from all over the transport sector and this has broadened my horizons to the art of the possible in terms of data science.
This is also reflected in the number of public consultations we run, from topics like a third runway at Heathrow airport to more niche areas like reducing emissions from machinery. We’ve issued 34 consultation documents this year alone making us one of the most prolific departments in government. Go DfT!
So, it’s no surprise that one of the first things we did with Data Science in DfT was to identify ways we could improve this process to make the jobs of our policy colleagues easier and save money for the taxpayer.
But where to begin? Adopting Agile, we started with the user need and spent time with policy colleagues to understand how they operated consultations and the pain points in the process. We quickly came up with two:
processing a large amount of responses via different media including letters, emails, documents and online survey responses
analysing and summarising what respondents are saying
Prototyping is easy...
Armed with this information we set about developing a prototype that could extract the text from the various media, and then process this into meaningful analysis that policy colleagues could interpret. Working with a small number of responses, we were able to pull a proof of concept that:
extracted the text from the documents
saved it to the cloud database
cleaned and processed the text
We built a topic model (https://en.wikipedia.org/wiki/Topic_model) to automatically assign the text into topics and visualised this interactively in a grand total of 2 weeks. Easy peasy! Or so we thought …
But solutions are hard!
Our first attempt at developing a data science application was anything but easy and we quickly found out that scaling from a prototype to a product was a real challenge. There was a myriad of additional issues we had to contend with. Architecture, hosting, deployment, security, error handling and authentication were all challenges we had to solve.
As such, it took us longer than anticipated to develop the first iteration of the application, and it didn’t even work well enough to use. As the saying goes, ‘experience is what you get instead of what you wanted’ – and we learned that lesson the hard way.
Humbled yet resolute, we looked back on development and identified areas where we could improve, both in terms of how we worked and the application’s functionality. We reflected, talked, planned and started work on the second iteration of the application.
Instead of trying to make everything perfect we adopted a ‘good enough’ mentality, but with added controls around checking and quality assurance and a more stringent approach to what features we implemented. This worked a lot better and the second iteration was smoother and quicker than the first.
As a result, we deployed the second iteration of the application to our cycle safety consultation. Did it work perfectly? No. Were there bugs and issues? Yes. Were our policy colleagues happy? Yes. Was it ‘good enough’? Yes. We could process and analyse several thousand responses within 10 hours, giving policy and ministers a high-level overview of the topics, supplemented by collated response documents. Normally, this would have taken weeks of painstaking and repetitive work.
It’s worth noting that the application isn’t a full substitute for the detailed analysis that policy colleagues conduct, but there is scope to improve this process and make it quicker and easier. Also, the application was applied to 2 other consultations at the same time and it didn’t work as well on these as it could have. If we were involved in the process earlier we could have influenced how data was collected from respondents for a better outcome. One of our next challenges is to work with policy to make sure their business processes work well with our application and vice versa.
It’s also very easy for us to dwell on what could have gone better. Given the number of government consultations, it’s no exaggeration to say that something like this could save the taxpayer millions of pounds through efficiency and more data driven decision making. As such, we’re going to be running the application on more consultations in future. We’ll continue to iterate, learn and improve with the aim of making those savings a reality.