$30 million of avoided capital allocation

Mosaic provided a regional conglomerate decision analysis for its coking coal project. We tested  operational vulnerability of its barge logistics plan. We stress-tested total barging day scenarios by backcasting the cumulative effects of deforestation and near-term climate change over mine life. Using deep learning procedures and asset scale climate modelling, we estimated the cumulative hydrographic impacts on saling days and the viability of the proposed investment.

Avoided $50 million in restoration liabilities

We provided an international palm grower with an evidence-based conservation planning and investment strategy. We stress tested scenarios to estimate the effects of climate change fire weather and population pressures on HCV forest resources. We translated the results with stakeholders securing buy-in for long term forest conservation.

Sizing up actual deforestation risk with deep learning

We helped investors better understand NDPE risk in a palm oil commodity landscape. Our workflow, driven by predictive algorithms tested and ranked forest resources vulnerable to loss and conversion across mill supply shed. Our solution provided an empirical basis to factor emerging ESG risk from forest conversion, fine tune risk management planning and action plans.

Undeclared red flags in mineral supply chains

Our remote diligence helped mining experts Britmindo assess actual governance at a low GAR coal property supplying the Sumatran power grid. Without ‘boots on the ground’, we enabled analysis of mine development rates, potential sterilization and restoration liabilities due to community activities over a 5 year period. Our hybrid detection methods, combining SAR radar and optical satellite data streams, cut through persistent clouds, smoke and haze to provide independent analysis of potential red flag issues.

Guiding toll road planning from space

We examined 40km toll alignment linking the Patimban deep sea port facility with a planned industrial city development. We provided the design team with design intelligence on surface waters between 1985-2018. Programming a classification algorithm – trained on 3 million satellite images – we rapidly identified optimal drainage across the entire alignment.

Identifying ground and structural movement up front

Mosaic provided a mega industrial city development with a rapid pre-construction health check including screening a third party toll road intersecting the development. Partnering with Terradue’s AI platform, we extracted and tested millions of ground surface movement points from SAR satelite radar imagery to identify areas of potential subsidence and structural movement on and off-site.

Enhancing traceability for palm oil

Mosaic developed a proprietary mill supply accounting method to enhance sourcing and engagement plans across a palm commodity landscape with at-risk forest and peat resources. Using machine learning workflows, we analyzed hundreds of roads and tracks making up local transport networks, calculated travel routing times to improve the client’s infromation on the actual supply shed risk.

Assembling risk intelligence on project communities

We provided Business Risk Indonesia with insights to inform community engagement and sharpen fire risk management plans for a mine facility. The client required time-series evidence of community footprints across the project landscape. We prepared spatial and temporal fire pattern and land conversion profiles that helped to identify key actors and ways to improve management of at-risk operational and exploration areas