More than 130 of you sent your ideas into the 2022-23 Public Service Data Challenge, and the judges then long-listed the seven most promising for development by interdisciplinary, cross-departmental project teams: these ideas are outlined below.
Our congratulations to those whose ideas were selected. And many thanks to everyone who submitted an idea: the quality of your proposals was very high, clearly demonstrating the expertise, inventiveness, enthusiasm and commitment of Canada’s federal workforce.
The project teams then researched their ideas – developing the evidence, data streams and partnerships required to demonstrate their value and viability. One team did not complete this process: in any innovative, experimental process, this is very much to be expected. At the Semi-Final on 24 January, having heard the remaining six teams’ pitches and put their questions, the judges selected four to go forwards to the Final on 21 June.
Learn which team triumphed at the Final on our homepage, and see the finalists’ ideas by clicking on ‘Learn more’ below.
Plastic waste and discarded ‘ghost’ fishing gear presents a serious threat to the health of our oceans – harming sea life, damaging delicate environments, and ultimately entering the human food chain. The Government of Canada is investing in clean-up operations, but these are extremely costly: an AI-assisted platform would help target this work, much improving its efficiency and effectiveness. Drawing on satellite imagery and sonar data, the tool would create heat maps of plastic and ‘ghost gear’ densities; it could also be further developed to track other pollution sources such as oil spills.
Idea submitted by Riham Elhabyan, Fisheries and Oceans Canada
Under its ‘proactive disclosure’ rules, the government publishes large amounts of data on federal spending; but the current interface is not intuitive and has limited functionality.
An interactive data visualisation tool drawing on contracts, grants and outcomes data would much improve its value to users, allowing them to compare datasets and view the data by – for example – program type, area of operations or correlation with improved outcomes. A secure area of the site could enable public servants to share information, advice and contact details, improving procurement operations across government.
Idea submitted by Idralyn Alarcon, Prairies Economic Development Canada
Ensuring that Canadians’ food supplies are safe and legal, the Canadian Food Inspection Agency carries out checks on imported goods – but how to decide which consignments to inspect? Staff currently scroll through long lists of shipments, using their experience to make judgements; but a Machine Learning tool could much improve the efficiency of this process. Fed data on which shipments have been inspected in the past, it could pull similar imports from the list in moments; and given information on where checks have revealed something amiss, it could better target inspections – averting the risk that non-compliant foods reach Canadian citizens, while strengthening deterrence. The tool could also be reconfigured for use in other compliance and enforcement services across government.
Idea submitted by Aaron Schamber, Canadian Food Inspection Agency
As noted above, this team did not proceed to the Semi-Final.
Crown-Indigenous Relations and Northern Affairs Canada (CIRNAC) offers a range of support programs for indigenous and isolated communities, seeking to allocate funds in ways that are transparent, impactful and equitable. These decisions could be improved by the development of a community needs index, presenting data on topics such as a community’s purchasing power, remoteness, social capital and vulnerability to climate change. And such an index would be valuable to actors across government, informing policymaking and service delivery in fields such as economic development, infrastructure provision and health care.
Idea submitted by Ioana Spinu, Crown-Indigenous Relations and Northern Affairs
Every year, departments spend millions of dollars carrying out consultations. If the results could be pooled in a data hub, we would quickly build up a huge dataset covering a wide range of topics – providing researchers with a broad and detailed picture of Canadians’ interests, situations and needs, and thus improving decision-making in policymaking and service delivery. Creating a hub could also assist departments to coordinate or merge their consultations, reducing duplication and addressing ‘consultation fatigue’.
Idea submitted by Thom Kearney, Privy Council Office
Through its AgPal website, Agriculture and Agri-Food Canada provides information on thousands of federal, provincial and territorial programmes and services for farmers and others involved in agriculture and food production. Currently, users must search by selecting categories from various lists; but such systems often struggle to meet people’s needs, omitting appropriate schemes while suggesting irrelevant ones. Today’s AI Natural Language Processing technologies could transform this search function, allowing users to ask questions and providing far more accurate and personalised responses. An AI-powered AgPal search function could thus help ensure that public funds reach their intended recipients, improving the targeting of services and the support offered to businesses.
Idea submitted by Jay Conte, Agriculture and Agri-Food Canada
When heavy rain is forecast, farmers in sensitive areas are asked not to apply fertilizers – which, carried into streams and rivers, can reduce oxygen levels and choke up waterways. This guidance is important, but it is a blunt instrument: farmers are dependent on public forecasts, and the advice takes no account of local geography or farming practices. Bringing together records on precipitation and water pollution, we could analyse how past rainfall has impacted water quality across the country. Then, linking this information to new weather forecasts, we could generate far more granular, accurate predictions of the local pollution risk – averting both pollution, and the issuing of alerts when the risk is low.
Idea submitted by Gurkanwal Arora, Environment and Climate Change Canada