5. Putting a pilot project plan to action

Formalizing a project team and working group

Here is a possible sequencing of scenarios regarding the project scope and organization:

Minimal viable initiative

This is a limited start with open data and may only have one individual working on it ideally with an internal working group for feedback. This initiative may be only data focused or policy focused.

Open data pilot project

This is the most common initiative for starting up an open data initiative. It typically has a project team, an internal working group and/or an Open Data (possibly Open Government) Steering Committee.

Open data program

A new municipal program requires assigned human resources and budget. This would typically follow the results of a pilot project and may draw personnel from that experience. The program resourcing is outside the scope of this initial documentation.

The makeup of the pilot project team should consider the following:

Project lead:

Often this is the individual who is closely involved with initial data offerings (typically GIS) and/or the lead for the proposed project. This individual is responsible for project planning, execution and reporting.

Project team:

There are many moving parts in municipalities’ data and information processes. Team members, depending on resources, could include: the municipalities Clerk/FOI Manager/Records Manager; IT Manager/Director or GIS Manager/Business Solutions Manager and Data Management/Business Intelligence analyst. This team is responsible for carrying out the pilot project per assigned tasks.

Working group:

Members for the working group could include some or all of the project team and have representation from other departments (e.g. Land Use Planning and Development; Economic Development; Engineering and Public Works; City/Town Managers Strategic Initiatives office; City/Town Solicitor and Communications). The working group is responsible for contributing to the development of processes, protocols and departmental data, if applicable, for the pilot project and providing representation for their department/divisional perspectives.

Steering committee:

This may not be necessary at this point or it could leverage an existing Steering Committee to report to. The role of a steering committee would be to provide project direction and coordinate with other corporate projects that could benefit from elements of open data.

Elements of this structure can be extracted for use in the open data project.

Confirming project resources

Resources committed to the project via the approvals process should be confirmed for the duration of the project. If at any time, resources become unavailable (e.g. assigned to another project) for a considerable period of time, notify the project sponsor and adjust the project schedule or acquire suitable replacement resources (possibly from within working group or external support resources).

Understanding data governance roles

Municipalities may vary in size and capacity to provide services; however, every municipality uses data on a daily basis for service delivery, operational process and longer term planning. For everyone involved in creating, maintaining, publishing and using data it is important to understand the different roles of staff related to data governance. Indeed, there can many questions related to data governance which can be a very complex management issue. For example, here some questions from a recent OKI blog on data governance:

Decision makers:

Who leads/asserts decision authority on open data in meetings, procedures, conduct, debate, voting and other issues?

Data holders:

Which organizations / government bodies manage and administer data?

Data producers:

Which organizations / government bodies produce what kind of public sector information?

Data quality assurance actors:

Who are the actors ensuring that produced data adhere to certain quality standards and does this conflict with their publication as open data?

Data gatekeepers/stewards:

Who controls open data publication?

Data Governance - Key roles

It’s clear that data governance is a complex topic. For purposes of the pilot project, identify these key roles:

Data custodian:

Those individuals (typically IT/IS department) who are responsible for providing data storage, data management, data archiving, data integrity, data security and data publishing for corporate data.

Data steward:

Those individuals within each department who are the creators, maintainers, quality assurers and distributors of business data.

Data users:

Those individuals who are responsible for using the data for analysis, visualization, report generation and identifying data quality and data integrity issues.

For small or medium size municipalities that do not have the capacity to resource these different positions, it remains a useful exercise to describe these different responsibilities and determine how they can be included in existing roles among staff.

Acknowledging open data as a cultural shift

Governments have been characterized in the past as being “closed”, risk averse and not as citizen centric as they could or should be. As we enter the realm of open government and open data, it is important to understand that this will be a process that is counter-intuitive to many government workers. Municipalities do not often have a data centric approach in their operations viewing data as a “necessary evil” as opposed to viewing data as a strategic corporate asset. As open data enables government to share public data with public to anyone who wishes to reuse it, it also breaks down silos within public administration. There is a paradigm shift at play: government is now responsible for the quality of publicly accessible data and no longer its proprietor. The long terms benefits of creating more responsive government institutions necessitates opening up government decision-making. This is both a challenge, in so far as it changes the status quo, and a tremendous opportunity.

It is likely that some resistance to change (i.e. from closed to open) will occur in most organizations. The other factor of being risk adverse will need to be addressed. The following links provide examples of perceived risks and their mitigation.

Confirming open data principles

The general democratic principles of transparency, accountability and being participatory are the foundation for the open government and open data movements. There have been a series of updated principles since 2008 and today the principles that are recognized internationally are provided by the International Open Data Charter. In Canada, the Federal government, Government of Ontario and recently Edmonton has become the first municipality to adopt the Charter. These Canadian governments have adopted the Charter by integrating its principles into their respective policies and open data strategies and frameworks.

The principles are:

1. Open by default
2. Timely and comprehensive
3. Accessible and usable
4. Comparable and interoperable
5. For improved governance and citizen engagement
6. For inclusive development and innovation.

The detailed descriptions of the principles may be found here. The review of these principles may seem a little intimidating but they become more easily understood in the context of the broader principles of transparency, accountability and being participatory.

It is also valuable to understand the keys to access and privacy as defined by the Ontario Information and Privacy Commissioner’s Office:

• Access by Design
• Privacy by Design

Developing the policy

Every municipality will have a variety of departmental and corporate policies. There may be templates to follow related to the development of the proposed policy. In developing an open data policy, it’s important to include the relevant people associated with data and information management. Based on being responsible for Freedom of Information requests and corporate Records Management, the sponsor for the policy might be best suited for the Office of the Clerk. The development would benefit from consultation with the IT/IS department, the City/Town Solicitor and departments that maintain significant data holdings.

There are a variety of open data policy guidelines with one of the most comprehensive being from the Sunlight Foundation and may be found here. These are quite specific and for a corporate policy document it would be beneficial to simplify the policy to address the following:

• Policy intent, open data background and goals for the organization;
• Principles of open data - International Open Data Charter;
• Governance, administration and roles within the open data program; and
• Community engagement of open data.

Assessing need to change existing policies

In most cases, the existing policies related to access to information and transparency complement open government and open data principles. All municipalities maintain a Fees Bylaw/Schedule that defines costs associated with a variety of products and services which can include fees for data (e.g. GIS data). Costs for data must be removed from the fees bylaw. Municipalities that have undertaken open data have removed the fee restriction to align with the principles of open data. It is important to always remember the needs of data users who may seek multiple data sets from different sources. The cost of administering data transfers, billing and collection of fees will be similar to fees and should therefore be eliminated for enhanced customer service and reduced operating costs.

Data lifecycle considerations

A data lifecycle approach reminds us that the creation and publishing of open data requires attention to additional tasks, such as use, archiving, and ongoing maintenance needs.

Current data governance may or may not have formal data governance rules around archiving data. This becomes especially important for people using open data “snapshots” to look for trends over time.

Similarly, the importance of keeping open data refreshed according to the existing data maintenance processes and timelines cannot be overstated - people expect open data to be current (within reason).

Another important aspect of open data management relates to the ability of exporting data from technology “business solutions” to ensure ease of extraction. This point needs to be considered for future technology acquisitions (i.e. not a closed system).

Data inventory

This task is an information management best practice but not always done as part of the pilot project. The documented conversation between municipality and community often goes like this:

City: What data would you like?”

Community: What data do you have?”

City: “Not sure, what data would you like?”

The data that you release should correspond to a specific demand from stakeholders that have an interest in accessing and reusing this data for their purposes. Ideally, undertaking at least a quick, preliminary inventory will help you with the prioritizing process in identifying high value data sets and having fruitful discussions with potential end users.

Prioritizing and vetting initial datasets

This is an extremely important series of tasks in the open data pilot project. The tasks may be summarized as follows:

• Undertake preliminary vetting of data re: privacy and other restrictions;
• Consult with internal and external stakeholders for “high value” datasets using a standard scoring system;
• Undertake a final assessment process to confirm suitability of the data for publishing;
• Prepare final list of initial datasets to be published.

Here is a list of the top 20 datasets that were used in the Open Knowledge Canada Local (City) Open Data Census 2016 (descriptions are from the census)

Public facilities

Location information about various Public facilities such as schools, parks, hospitals, daycare, pools, rinks, city service counters, water fountains, bathrooms etc.

Election results

Results by constituency / district for all major local electoral contests in this municipality

Road construction

Planned road construction with start-end/duration

Natural resources

Natural assets including tree inventories, Environmentally Sensitive Areas, Areas of Scientific Interest, endangered areas, etc.

Building permits

Building / construction permits

Zoning (GIS data)

The type of development allowed on parcels of land (e.g. residential, mixed-use, commercial, industrial, special use)

Service requests

Non-Emergency service requests to municipal authorities, for example regarding potholes, graffiti, etc. (311, D115 etc.). Data should be at granular (per request) level.

Annual budget

Municipal budget at a high level (e.g. spending by sector, department etc.). This category is about budgets which are plans for expenditure (not actual expenditure in the past).

Transit schedules

Timetables (schedules) of all municipally run or commissioned transit services (buses, subway, rail tram etc.). Locations of stops would also be good (as geodata).

Business permits

Business Permits / Licenses

Business listing

Key information for businesses in the municipal area, such as name, address, contact information, business type.

Food safety inspections

Outcomes of food safety inspections of restaurants and other similar providers of food to the public

Expenditures (detailed)

Records of actual (past) municipal spending at a detailed transactional level, for example, at the level of month to month expenditure on specific items (usually this means individual records of spending amounts at a fairly granular level - e.g. $5-50k rather than at the $1m+ level). (Note: a database of contracts awarded or similar is not considered sufficient. This data category refers to detailed ongoing data on actual expenditure)

Transit: real-time

Real-time information about major municipal-run or commissioned transit services (buses, subway, rail, tram etc.). Real-time transit information means things like the location of actual services (individual buses and trains, etc.) or updated stop times that reflect service locations reported through sensors (such as GPS).

Crime statistics

Data on municipal crime, preferably at a reasonably disaggregated level (best would be exact date, location and type but per day per street or postal code would be acceptable)

Procurement contracts

Per contract information on municipal contracts including amount, awardee (name, address), date awarded, date completed, penalties, etc.

Traffic accidents

Statistics on road traffic accidents including time and location

Re-zoning permit application

Application for permit to rezone a parcel of land

Campaign finance contributions

Amount contributed to each candidate and by whom.

Lobbyist activity

Actions of named registered lobbyists.

Some of these datasets may not be within your organization’s jurisdiction but does indicate the type of data that has been identified of being of interest to community.

Vetting the data inventory

Based on your current inventory of data sets, there needs to be a vetting process undertaken with appropriate staff including data custodians, information/privacy officer, City/Town Clerk, Solicitor and Open Data Pilot Project Lead.

Open Data - Dataset identification, review and publication process - Text version

This Figure depicts Niagara Region’s four –step use case approach on Open Data identification, review and publication. The first step on identifying data is completed within all departments at Niagara Region. The second step on assessing data is completed by staff from across organizations with data quality review being completed by data expert; departmental approval being completed by data expert with endorsement from departmental management; privacy and security review being completed by information management, legal and copyright review being completed by legal services (**optional), and communications review being completed by communications. The third step of preparing data is completed by IT staff in Corporate Services and the final step of publishing data is completed by IT and Communications staff

The right hand side of the Figure provides a detailed description of the four steps in a text format.

Step 1 is completed within all departments at Niagara Region. There may be dedicated staff resources (Data Experts ) that deal specifically with Open Data and identifying datasets for possible inclusion in the Open Data Catalogue

Step 2 is completed by staff from across the organization. Please refer to the following breakdown:

1. Data Quality Review – Completed by Data Expert
2. Departmental Approval – Completed by Data Expert with Endorsement from Department Management
3. Privacy and Security Review – Completed by Information Management
4. Legal and Copyright Review – Completed by Legal Services (**Optional)
5. 5)Communications Review – Completed by Communications

Step 3 is completed by IT staff in Corporate Services.

Step 4 is completed by IT and Communications staff.

Prioritizing the datasets

For the pilot project, determine the most desirable datasets for release - the “high quality” datasets.

There are a variety of parameters that can be used for identifying “high value” datasets for the pilot project. It’s also important to ask the question: high value for whom? As an example, here is a list from the City of Vancouver’s “Open Data Policy and Implementation Framework” document:

The prior example parameters for prioritizing datasets can be further assessed based on a scoring system.

Tool: City of Vancouver data prioritization

Open data standards

There are a large number of data standards that exist through organizations such as ISO, OGC, IEEE and national organizations such as the Canadian General Standards Board. For open data, we’ll focus on three areas:

• Open Data End User License;
• Open Data formats; and
• Open Data Metadata.

Open data format standards

One of the key attributes of open data is that it be machine readable. Open Knowledge International defines machine readable as “ Data in a data format that can be automatically read and processed by a computer, such as CSV, JSON, XML, etc. Machine-readable data must be structured data .” A more detailed discussion on machine readable data can be found here: http://opendatahandbook.org/glossary/en/terms/machine-readable/ and a glossary of data format descriptions can be found here City of Guelph Glossary

For data formats, most users will indicate that they have preferred data formats that allow ease of use of the data. Based on current best practices, the following data formats should be available for open data: CSV, CSV/XLS; TXT; XML; JSON; and for geospatial data – Shp, KML/KMZ, geoJSON and Mr SID (raster). This list is expected to evolve as the user community expands.

For purposes of the pilot project, it is important to have data which is machine readable to facilitate ease of use of the data by end users (e.g. academic researchers, apps developers and citizens).

Another consideration related to access to information is the development of Application Programming Interface which supports innovation through programed access to your data. This does not need to be part of your pilot project other than possible discussion with the user community on the potential benefits to them. A short discussion on this is provided below.

Open data metadata standards

The importance of metadata (data about data) cannot be overstated. The end user requires some qualitative information about the dataset you have provided. Metadata (data about data) is essential to provide information related to each of the data sets that are made open. Metadata standards have existed for some time and it makes sense to use what already exists.

For non-spatial data, it is recommended that the 15 elements of the Dublin Core Metadata standard be used with the addition of a parameter for data quality. For geospatial data, it is recommended that the ISO TC-211 19115 NAP (North American profile) Metadata standard be used. Metadata standards are typically included with software you may be using for publishing the data and in fact can often be populated at the time of data creation. This especially applies to creation of GIS/geospatial data which is often considered the “low hanging fruit” for starting open data due to the structured data and often built in quality assurance. There are more standards that can be found in the additional resources section and the goal of having consistent data standards throughout Canada continues to evolve.

Data publishing options

How you publish your data is the primary technology decision that needs to be made in the pilot project. Typically, the starting point is creating an open data “catalogue” (location to access the open data). The following are the four basic options available for publishing:

• Make data available via a simple HTML page;
• Develop your own catalogue publishing software;
• Leverage existing data publishing software; or
• Pay a third party to upload the data and manage your open data site.

From the above, the first choice (HTML) is a logical approach if you have scant resources and a critical timeframe to meet. The second choice of developing your own makes no sense - there’s existing software available (proprietary and open source). The third choice is typically the one chosen by most municipalities looking to create a data catalog. This can be accommodated by use of open source software such as CKAN (can be customized and is used by Federal, Provincial and municipal governments) or proprietary software such as ArcGIS Online Open Data portal.

Here are some examples of the variety of open data publishing websites:

• HTML based open data catalogue. City of London, ON
  Open Data catalogue.
• Open Source CKAN based with customization. City of Surrey, BC
  Open Data Catalogue
• Proprietary ESRI ArcGIS Online. Halifax Regional Municipality, NS
  Open Data Catalogue
• Socrata 3rd party hosting service. City of Edmonton, AB
  Open Data Portal

Data publishing - alpha and beta versions

Once you’ve assessed your publishing technology options, it’s time to build the website. There are numerous examples from existing websites to provide insight into the content you should consider. Here’s a quick checklist of content requirements:

• Introduction to Open Data;
• Open Data licence;
• Open Data with metadata;
• Full-text search (to assist in finding data of interest);
• Data Request form;
• Contact information.

Many sites are now incorporating analytics and visualization; however, this is not considered necessary for a pilot project.

Publishing Processes

For open data publishing to run efficiently, automate processes as much as possible. The automation may require some additional technology resources or some custom development of in-house tools. Consider the actual publishing to be available as a decentralized service where individual departments could upload data on their own. The following are some example processes.

Open data metrics

Define some Key Performance Indicators (KPIs) to illustrate the impact of the project. The following is a list of potential metrics that can provide some indication of the impact of the pilot project:

• Initial Open Data
  • Requests for additional data
  • Requests for data transformation into other formats
  • Requests for additional metadata
• Use of the Data.
  • Online survey regarding ease of use, time savings and value of service
  • Meetups with users to discuss value of the data, what should be next datasets released
  • Website analytics - time on site, returning users …
• Impact
  • Reduction in data requests to internal departments
  • Number of application development projects
  • Number of community organized open data related events

There are a number of well-respected non-profits that undertake open data metrics including:

• Open Knowledge International : Global Open Data Index
• OK Canada: Local Open Data Census;
• World Wide Web Foundation: Open Data Barometer

Communications plan

Each municipality has someone assigned to look after communications with either internal staff resources or outside contracts for communications support. The general goal for the communications plan should be to create awareness of what open data is and the benefits that it can provide for citizens, social and private sector stakeholders and others. As a general rule, it is best not to overdo your communications on the day of your launch so that it is not perceived as a public relations exercise. It’s great to claim credit for a new initiative, make sure that you have a follow-up plan after your big announcement. Releasing your valued data set is the end of the beginning. Your messaging should also reflect the values of collaboration at the core of the open data community. There are opportunities for community events, such as International Open Data Day, to provide an excellent communications opportunity.

The following are a few suggestions related to open data:

• News/Blogs – need ongoing open data news to keep site fresh.
• Social Media – consider Twitter, Facebook and YouTube for exposure.
• Feedback/Ideation reporting – need commitment to respond to feedback and ideation suggestions from staff and external stakeholders.

Prior to the launch, there are a number of communications related tasks that should be undertaken including:

• Workshop/Information session/Lunch & Learn for staff;
• Invitations to Launch for community (citizens and businesses) and politicians (often done as a Council presentation);
• Social Media material (e.g. tweets, Facebook); and
• Press Release for launch.

Leveraging alpha open data catalog for community engagement

The pre-public release of the alpha website is an ideal opportunity to build out your community. In many cases, there has been an informal community network which might now become more formal similar to these examples. This is an opportunity to bring together “birds of a feather” interest groups in innovation, technologies and building a better community.

Keeping the conversation going

It’s important that people feel engaged in the project and that requires ongoing communications. Short meetings with interested staff in conjunction with working group department leads will lead to enhanced opportunities for cooperation and collaboration in the opening up of data in the municipality.

The community groups may have a variety of interests in the pilot project from data formats, to use of data via the open data license to developing applications with the data to looking for data that assist with community issues. The short and longer term success of open data initiatives can be traced back to the quality of the relationship between the municipality and its community - open data is a great opportunity to build out the community.