Open Data Strategy for Ireland Submission 2014

As part of the public consultation process in respect of the development of a national Open Data Strategy for Ireland, Dublinked made a submission to the Department of Public Expenditure and Reform in September 2014.

The Department of Public Expenditure and Reform is developing a national Open Data strategy for Ireland in line with commitments contained within Ireland’s Open Government Partnership Action Plan 

Launched on 22nd July 2014, the national Open Data portal data.gov.ie  details 415 datasets from 45 public bodies. As part of the preparatory work for the strategy, Insight Galway have produced the following reports –

  • Best Practice Handbook – this is the core document which draws together existing best practice standards for the publication and re-use of Open Data to assist in establishing best practice standards in Ireland
  • Data Audit Report – reports on an audit of the Irish public sector datasets available online and aligns the findings with the ‘common high-value datasets’ based on the G8 Open Data Charter categories
  • Roadmap – sets out a detailed 3 year plan for the objectives, structure and actions necessary to advance publication of national and regional Open Data for maximum impact
  • Evaluation Framework – to assess the current state of readiness for Open Data in Ireland; to monitor and assess the ongoing progress of the initiative, and to measure the actual economic, social and political impact of the initiative.

Further information can be found on the Department’s website at http://www.per.gov.ie/open-data/ 

The following is Dublinked’s submission –

Background

Dublinked[1] is an internationally recognised regional innovation network that enables collaboration between industry, public sector and academia to address data-driven challenges and promote economic activity.   Dublinked was launched in October 2011 and is a partnership of the four Dublin Local Authorities and the National University of Ireland Maynooth with technology provided by IBM.

At the heart of Dublinked is the vision that the Dublin region can be a testbed for innovations that have commercial potential. Dublinked has emerged as significant network for researchers, developers, public bodies, local authorities, data providers, small medium enterprises and multinational organisations to connect, share knowledge, interact with regional data and pilot solutions to regional challenges.

The innovation network is a single-point-of-contact for new companies and users who wish to engage with the public sector for data requests and project proposals. Through facilitation of networking opportunities and hosting of thematic events, Dublinked brings together interested parties from different sectors and domain experts to share ideas and challenges, develop solutions for Region and identify opportunities for collaboration. In developing new relationships with developers, start-up community and entrepreneurs, local authorities, public bodies and agencies by sharing data, receive direct feedback about the kind of data they need to build to improve services.

 

Datasets

The range and number of datasets published on Dublinked has increased significantly, to include in excess of 300 datasets, since the initiative was launched in 2011. Data providers have expanded beyond the Local Authorities to include national bodies, such as National Transport Authority, the Railway Procurement Agency and include datasets from private sector e.g. Dublin City Business Improvement Districts pedestrian footfall data.

One of the novel aspects of Dublinked is pooling of confidential or sensitive data between multiple partners, through both an ‘Open Zone’ and a ‘Research Zone’. The generation of this common pool of research-level data is unique to Dublinked.

Over the three years Dublinked has made significant progress in achieving its vision, aims and objective:

Dublinked is an internationally recognised innovation network that drives economic development through sharing open data and connecting people through events, workshops and social engagement.

A strong motivation for the Dublinked partnership is the conviction that the Dublin Region can be a test-bed for data-driven innovations that have commercial potential. In its pilot phase the project demonstrated its value as a portal for sharing knowledge, experiences and data, hence becoming a pivotal element of the local ecosystem for innovation and entrepreneurship.

Dublinked has built an international reputation with users from over 128 countries for promoting data-enabled entrepreneurship, fostering an environment conducive for both large and small companies to use open-data in the context of the Dublin region.

To illustrate – the project was recently cited in an international comparative study on “Innovative Urban Development Strategies for Sustainable Competitiveness”, by Euricur and Price Waterhouse Coopers[2], ‘as starting to contribute to improve Dublin’s structural assets or “capitals”, towards a new stage of more sustainable competitiveness’ carried out by.

As the first open-data initiative to take an economic development perspective, it is viewed as an exemplar by international institutions in the field and other cities.

Dublinked has emerged as a significant network for researchers, developers, public bodies, local authorities, data providers, small medium enterprises and multinational organisations to connect, share knowledge, interact with regional data and pilot solutions to regional challenges.

Data providers have expanded beyond the Local Authorities to include national bodies, such as the National Transport Authority,  the Railway Procurement Agency and include datasets from private sector e.g. Dublin City Business Improvement Districts pedestrian footfall data.

Dublinked reinforces Dublin’s growing reputation as a tech sector destination, particularly in the rapidly growing area of open-data and smart city initiatives.

This growth is demonstrated by related developments in this area, for example:  Digital Leadership Forum, Digital Masterplan for Dublin, CityWatch, Activating Dublin, Dublin Dashboard and the growth of an active open data community with the recent launch of Code for Ireland. All of these initiatives involve citizens co-creating the city region and suggesting innovations.

Dublinked helps the Dublin local authorities to fulfil their open data national policy obligations. New national policy frameworks, such as the eGovernment Strategy (April 2012), the Public Service Reform Plan, the Open Government Partnership and the Action Plan for Jobs reiterate the government’s commitment to initiatives utilising open data.

In particular, these frameworks focus on performance objectives for the public sector such as enhanced information sharing and improved knowledge management practices. Dublinked provides a cost-effective mechanism for managing external requests for data, and release, thus avoiding duplication and improving user engagement.

Dublinked builds private-public institutional collaboration. The project has become a forum where multiple stakeholders can effortlessly contribute knowledge, actively engage in supporting innovation and further promote Dublin’s economic development.

As a bridge between the different regional initiatives Dublinked contributes to greater cross-sectoral research collaboration between educational institutions, entrepreneurs and large companies such as IBM, Intel, Siemens and Nokia.  The project also provides a vehicle for Horizon 2020 proposals and information sharing across other European projects.

Dublinked helps to make the services of Dublin’s local authorities more customer-centred. The project provides a network of new relationships with developers, start-ups and entrepreneurs whereby the Dublin local authorities receive direct feedback about the kind of data they need to build to improve services.

“It was very useful for us to hear feedback from the developers about how they consume the data so we can include new procurement criteria to ensure our data is ‘fit for purpose’ when tendering for new systems “  Intelligent Transport Systems Engineer, Dublin City Council

“Because of Dublinked we were able to see some of our own information for the first time” Information Systems, Dublin City Council

Dublinked Innovation Network – Inspirational Reuse

At the heart of Dublinked is the vision that the Dublin region can be a testbed for innovations that have commercial potential. By releasing data and sharing experiences and expertise, we have shown that opportunities can be identified and innovation solutions developed, for example Dublin was used as a pilot for parking application ‘ParkYa’ and a planning webservice ‘Building Eye’, which is being scaled up in the United States.

The release of perceived ‘high-value’ data such as Real Time Passenger Information and Traffic Flow data has proven very popular with significant access to and reuse of this type of data.

Dublinked does not require members or users to list their reuse-developments, but does actively encourage the sharing of reuse stories through the Dublinked Inspirations page and Innovation Events. Some examples of Dublinked reuse are detailed below;

Real Time Passenger Information Projects– Dublinked provides access to Real Time Passenger Information (RTPI) API. To date there are over 38 registered users of the RTPI API, following a survey 14% have developed application incorporating this data, a further 62% or working with the RTPI to develop applications. 22% of users are utilising the data to communicate insight such as product positioning or service delivery.

In the month of December 2013 the RTPI API recorded 235K hits, with 103 different IP addresses pulling data.

A team from IADT were winners of SAP InnoJam 15 hour challenge with an app called “Go Simply”. The team used RTPI data from Dublinked to build their which report variability of travel durations on public transportations during the course of the day.

Visualisations; There has been an increase in developers using Dublinked to gain better understanding and communicate insight by creating data visualisations. Presently there are 10 visualisation based on Dublinked dataset including; 1) Analysing Ambulance & Fire Brigade call out times across the Dublin region 2) Creation of real time visualisation based on Trips traffic flow feed illustrating travel times 3) Graphic representation of fixyourstreet request through Dublinked data feed.

Applications: There are in excess of 20 SMEs and application using Dublinked. Over 80% of applications (detailed on Dublinked Inspirations page) relate to transport including ParkYa, HitTheRoad, Moovit, NearbyBus and Transport Times. In addition to producing applications SMEs are also utilising Dublinked to improve product offering or service placement such as AppleGreen, Deloitte, MetroHerald and RateMyArea.

Research: Dublinked has been used as vehicle to share research and build regional connectivity for the purpose of research funding. In 2012 Dublin City Traffic Department released sample data to aid in application for EU funding.  Dublinked was also highlighted by IBM when awarded second place in the International Semantic Challenge for the creation of SPUD (Semantic Processing of Urban Data), which is described as “a technology demonstrator for cataloguing, exploring, transforming and understanding urban information developed in the Smarter Cities Technology Centre of IBM Research.”

 

Open Data Strategy

Dublinked welcomes the development of an Open Data Strategy for Ireland, and believes there is an opportunity to include recommendations that would further develop the work that we have been engaged in relation to regional data driven innovation.

The following are Dublinked’s recommendations for consideration in developed of Ireland’s Open Data Strategy:

Recommendations

  1. Clarity is needed on what publishing data on the government portal means?

There is an ambiguity in the document about publishing open data and the role of the national portal. In some places it refers to passing metadata to the portal and in other cases it implies the data portal hosting the data. This needs clarification. We would recommend a federated approach where metadata is provided but data does not need to be hosted by the portal.

  1. Emphasise data quantity over quality.

The recommendations of the various reports is targeted at achieving quality of the published data rather than the variety and quantity of the published data. Best practice is to initially aim for quantity in whatever form in order to develop momentum and a culture of release, as opposed to overly focussing on achieving the highest quality open data release. Once that momentum and culture of data release has been established, then the emphasis can shift to focussing on quality.

  1. There is a need for greater education and expertise for data providers, especially on data protection compliance

The report does not focus on the cultural and educational issues in relationship to open data, particularly the risks to the authorised data owner and the data generators. There are a number of subtle issues with respect to data protection, commercial sensitivities and operational issues that need to be addressed at different levels of each organisation. For example the challenges in achieving the appropriate level of anonymisation is not a solved problem, and guaranteeing its achievement in an automated fashion is very difficult. This has caused Dublinked ongoing problems which discourage openness.

  1. A recommendation on the supply of data from contractees back to the government should be included.

There have been a number of situations recently, here and abroad, where outsourced functions have resulted in contractual constraints on the data being released, irrespective of the wishes of the public sector data owner. A recommendation should be included that any public sector contracts should require operational and statistical data to be returned to the public authorities for dissemination.

 

  1. That metadata should be exported in accessible format from each data site and not be the sole responsibility of a single site for search and discovery

Under the recommendations for 10.2.4, it is recommended that all metadata should be provided to a single central location for search and discovery. Instead of this centralised process, it should be required that all data repositories should export their metadata in an open-data format (XML, JSON) that would be available to both data.gov.ie but also any other portal or operation that wishes to utilise the data and develop their own search and discovery tools. We agree that there should be a central portal, but recommend that it should not be an exclusive option. In practice, a federated platform that allows for integration of existing and new datastores into a discovery framework would be the most robust solution. Following this approach would allow for a more creative and flexible development of portals, optimised for different uses, and be more compatible with the spirit of open data.

  1. Dependency on the 5 star data rating

The 5 star data rating scheme being proposed by the best practice report has encountered a number of issues and we believe that the recommendations have built in an over-dependency on complying with the formulaic nature of 5 stars with metrics on 5 stars without looking at the practicality and best use scenarios. For example, in practice CSV is an ill-defined format which by itself is not useful. There are also many other datasets that come in proprietary formats which are highly valuable but are simply not replaceable by CSV or some of the other formats being proposed. Thus there is a risk of building in an expectation that we can achieve the impossible and often the undesirable. The 5 star rating scheme is useful as a reference but an over-dependence on the scheme, and the requirement of hitting 5 stars could be an expensive and unnecessary distraction from delivering useful open data in the initial period of time.

  1. Inclusion of data from Local Government and other Agencies

The pilot data.gov.ie site does not include data from existing Open Data initiatives including Dublinked. There are currently approximately 300 datasets available on the Dublinked website which could be included in the data.gov.ie catalogue. These include datasets from other agencies including the high value Real Time Passenger Information from the National Transport Authority. Future development of data.gov.ie should include data from Local Government, other agencies and other Open Data initiatives.

 

[1] Dublinked http://www.dublinked.ie

[2]Innovative Urban Development Strategies for Sustainable Competitiveness – Dublinked Report’, Euricur & PWC (2011)

Leave a Reply

Your email address will not be published. Required fields are marked *