Update on Timber-Paparazzis: Progress and Learnings

It has been a while since we last updated our business models for impact, nevertheless we worked on several projects and have some exciting news and learnings about the “paparazzi” business model to share:

The idea got funded

In partnership with the NGO Fairventures Worldwide we applied for the Google Impact Challenge Germany. After several assessment rounds and a final public voting, we won the public vote (with around 70’000 votes) and received an incredible 500’000€ cash prize to realize the proposed business model! A big thanks to everyone who voted! Congratulations to the team and a shout out to everyone who wants to contribute to making it a success!

Some learnings on the way

Shortly after the prize, we flew to Central Kalimantan on Borneo to discuss optimal ways forward with our team and lead users, from farmers to timber processors and local governments. During the stay we could already work on some of our key hypothesis.

Bridge17 with locals and project staff in the field

The good news is, the core technology seems to work. As all Bridge17 business models focus on a digital core for its business models, in that case an AI that detects the diameter at breast height (a standard term in forestry) is core of the solution. The team ran a first test by taking roughly 10k pictures of one tree species in the field and trained a neural network on a subset of this data. Already with this quick test, the team managed an accuracy of roughly 90% with this small data set!

Paparazzi taking pictures

On the other hand, as typically with business model design, the initial idea won’t hardly be the final business model. Spending time in Central Kalimantan and Java in Indonesia, talking to farmers, officials and other stakeholders, it became clearer that the farmer’s role – as being both user and customer – needs some rethinking. Most smallholder farmers, who own some land, are often not directly relying on the income from that land as first option – they have other main jobs or income sources such as other crops. Thus, both the attention to actually perform the monitoring as well the potential economic benefits don’t count as much as initially assumed from us. On the other hand, the technical solution and its potential value seems to be of great interest for players who already perform monitoring or plan to digitize their monitoring tasks in the near future.

Both hypothesis will be further tested with a first functioning prototype.

The way forward

With the received funding we are currently working on different milestones:

–          Phase 1: building the technical solution including a native App or Progressive web app and training the neural network further

–          Phase 2: rolling out the solution in Central Kalimantan to test the usability, potentials for errors and adapt the solution

–          Phase 3: rolling out the solution to farmers in Indonesia and Uganda, including proposed affiliators and evaluate new potential customers

Leave a Reply

Your email address will not be published. Required fields are marked *