Group items tagged

The device is a small blue box that measure the rate of temperature rise rather than detecting smoke which helps in reducing false alarms.[5] The network of Lumkani devices in a specific slum uses radio frequency to send text messages and notify people in cases of emergencies. An alarm will be activated in all houses within a 60 metres (200 ft) radius if the initial alarm is not disarmed within 30 seconds.[6][7]

Funded through a government innovation agency and a successful crowdsourcing campaign, the start-up has also won several awards and grants, including best start-up in 2014's Global Innovation through Science & Technology competition.Last year, it was chosen by the International Red Cross as a provider for its Fire Sensors initiative, a project that has distributed 900 of the devices along with extensive market research.Julie Arrighi, Innovation Advisor at the American Red Cross, said that using networked heat sensors like Lumkani is crucial to mitigating the growing risk of slum fires around the world.

The device detects fires by sensing for a rate of rise of temperature, rather than smoke, which would cause many false alarms. In the event of a fire, the detectors communicate to one another by means of radio frequency (RF) to create a community wide alert – the alert is transmitted to neighbouring detectors within a 60m radius of the fire, giving the community at large the necessary advanced warning to deal with the fire before it becomes unmanageable.

Machine learning is a critical component of this process, as it allows us to enhance our predictive ability to better identify these qualifying customers. As a result, we are making significant strides towards our goal of creating the world’s leanest and lowest-cost banking infrastructure. This helps our bank partners offer loans at the lowest possible rate, and it helps our customers to achieve higher-yield savings. We’ve proven that our model works. Twelve million customers in Africa and Asia have saved or borrowed on the JUMO platform, and roughly 70% of these customers are micro and small businesses. We’re incredibly proud of our progress so far – but we know there’s more we can do.

To test our approach, we looked at Ghana, running new and advanced machine learning algorithms on signals associated with a pool of lending applicants. The goal was to identify combinations of signals — from income size and deposit frequency to mobile phone battery life and mobile wallet transaction patterns — that would serve as more accurate indicators of likely fraud than existing models.