The Emergency WA Hackathon 2023

1. How can we provide citizens with hearing and vision impairment, better emergency interactions?

WA community members who suffer from partial or complete hearing and vision impairment cannot easily interact or keep up to date with warnings information being issued on Emergency WA. Warnings information can be issued at a high frequency, in some cases every 20-30minutes. DFES seeks to develop better communication methods for WA community members who suffer from partial or complete hearing and vision impairment.

2. When first responders attend an incident they need to relay critical information to Coordinators of Public Information. How can we prompt this conversation to make sure the most critical information makes it through first?

DFES Coordinators of Public information (CPI) receive telephone calls from emergency response teams requesting warnings. Experienced CPIs have gained considerable knowledge and experience in understanding the caller and asking clarifying questions to determine the correct warning level to issue. Due to the increasing frequency and severity of emergencies in Western Australia, DFES has increased the size of the CPI team and wants to ensure a standard experience is provided to all emergency response teams requesting a warning, especially during high stress situations (multiple concurrent calls) or when telephone call quality is poor.

3. How can we automatically generate information for people who are affected by an ongoing incident?

Alerts and warnings posted to the Emergency WA website contain critical information relevant to people in specific locations. People impacted by emergencies are often highly stressed, making it difficult to read through lengthy warning information to make quick decisions about what to do. Community research and feedback has told us people increasingly want to receive information that is personalised, rather than a ‘one size fits all’ approach. Therefore, DFES seeks to deliver highly personalised warnings to the WA community so that any individual receiving a message understands what they need to do, according to their own location and circumstances, including the safest route to evacuate (where required) These directions could then auto-load to their in-car navigation where available. The WA community access Emergency WA across a range of diverse communication channels including but not limited to: Emergency WA website, 13 DFES Telephone Information Line, Facebook, Twitter, Desktop and Mobile. In future these channels may expand to include In-Home devices, Wearables, In-Car apps, SmartTVs amongst others. During an active emergency such as a bushfire or cyclone, people affected often have questions spanning how to prepare, where to evacuate to, how to get there, what to take, etc. This information is available throughout the Emergency WA and DFES Corporate websites but is not always easy to find, and much of it is not provided in a format that is easy to understand during times of stress. DFES is seeking to develop a capability for the WA community to ask questions and be self-served information in an easily understandable format across any communication channel.

4. How can we detect and automatically translate emergency notifications in to a person’s native language?

The Emergency WA website contains alerts and warnings information in English only. Recent Census data (Census 2021) indicated 18.7% of of Western Australians reported speaking a language other than English. Of these people, 12.1% (approximately 33,000 people) reported as having low English proficiency. This was lower than in 2016 (when it rated at 14%). DFES seeks to deliver warnings to people who speak languages other than English, through real-time translation. Ideally, the translation tool/engine will have the ability to learn and improve based on user feedback. When mapping warning areas, DFES would like to know immediately what languages are spoken within a mapped warning area. This assists the incident response in activating translators and community representatives with specific skills and experience.

5. Can we automatically generate appropriate, contextual, warnings based upon a person’s location?

When mapping warning areas, DFES would like to know immediately what suburbs, towns, Local Government Authorities (LGAs) exist within the warning area boundary and the tool suggest an appropriate location to insert into the warning headline. This term would need to scale to use appropriate locations based on the size of the warning area. For example, a warning area could encompass a small area such as one suburb. In this case, the headline could read: Bushfire Emergency Warning – Take Shelter Now – Success. However, if the warning area spanned half of Western Australia, determining the location would be extremely difficult as suburbs, towns, LGAs would be ineffective.

6. Can we automatically generate warnings based upon trends and forecasts from disparate data?

DFES operators manually create community warnings using data from multiple sources such as the Bureau of Meteorology (BoM). For hazards that have greater lead time, such as severe weather, cyclones and floods there is the potential to fully automate the core warning message creation and publication to provide faster warnings to the community. Trained human operators can then be used to review warnings and provide quality assurance. For more time critical hazards, such as bushfire, data sources such as Computer Aided Dispatch data could be used to trigger automatic warnings combined with data from satellite imagery and bushfire risk plans.

7. Emergencies and incidents can occur anywhere and at any time. With the full set of data from DFES, can you identify what locations are most likely to be impacted by different hazards over the next year?