The Challenge
Kerbside dumping is an ongoing problem for local council waste management teams. Cleaning up dumped rubbish can cost councils up to $750,000 a year, and dumping also damages infrastructure and requires environmental restoration.
The council was frequently affected by dumped rubbish, which resulted in:
- Regularly scheduled waste management trucks being unable to collect the rubbish piles and mattresses
- Manual collection services being dispatched reactively rather than on optimised routes and loads
- A reliance on rangers and residents to report rubbish piles and mattresses, which could take weeks
The Solution
Blackbook AI built an annotated dataset using historical data from smart-enabled trucks and in-cabin triggers.
Using GNSS-linked camera footage from the garbage trucks, a computer vision model detects rubbish piles and mattresses and pinpoints their location. An API then sends the council the street addresses of locations that need a kerbside pick-up.
The Outcomes
The council was able to unlock the value of its raw camera data to drive decision-making.
Key outcomes included:
- Significant cost savings from route and load optimisation of dispatch services
- Faster response times for pick-ups
- Targeted campaigns and response monitoring
- Improved coverage through existing garbage truck schedules, with less reliance on rangers and residents
FAQs
How does the system detect kerbside dumping?
Why use garbage truck cameras?
Does this need new hardware on our fleet?
