Sustainability gets a champion in geospatial data analytics

A recent article in SpaceNews captures beautifully the use of Geospatial data analytics in the Sustainability / Environment element in ESG. We have been active in the broader data analytics sector and it's quite fascinating to see how this emerging stream covering geospatial data has the potential to make the physical world we live in as efficient as the digital one. Here we focus on one specific use case which is Supply Chain Traceability and related investments.

a. Supply Chain Traceability is a key component of sustainable sourcing, which has recently been driving business strategies. Companies are under constant pressure for consumers, regulators, NGOs and stakeholders to disclose more information on their supply chains. A company might have tens of thousands of Tier 2 or Tier 3 supplier relationships they don’t really understand. Associated reputational costs, fines for not meeting reporting requirements or possible ESG related issues could be quite high on the brands.

Harvard Business Review has built an interesting two-dimensional matrix framework to plot the landscape and track the evolution of traceability. There are already a few firms that could, in the medium term, potentially move to the top quadrant that translates to full disclosure at the raw material level. Link to the article.

b. Reimagining supply chain transparency using technology - Early adopters are using cutting technologies to gain first-mile visibility. Best is to illustrate this with Unilever’s collaboration with Orbital Insight that is piloting supply chain traceability for their palm oil sourcing in Southeast Asia. Unilever uses a number of raw materials and palm oil was an immediate choice to begin with, given its importance and the impact it has on natural resources if it’s not sustainably sourced.

• Satellite technology is absolutely important while measuring our impact on the environment. Some of the ESG data can only be obtained from a vantage point that satellites enjoy. But images alone can’t prevent deforestation.

• The current techniques using satellite images are very rudimentary - simply assume a fixed radius surrounding production areas to drive supplies and analyses this catchment area to determine impact. While this is better than manual surveys, it’s not accurate, real-time, efficient, and lacks granularity.

• As you would imagine, sourcing in agriculture is complex - crops, before being processed, can be harvested from many different areas of land, belonging to numerous farmers, and mixed with raw material from other farms several times over before even reaching the mill.

These teams have combined GPS data, satellite imagery, proprietary AI algorithms and scalable data science to overcome all these challenges and provide a sophisticated, powerful traceability platform that can be scaled across raw materials and regions.

Analysis of aggregated and anonymized GPS data helps spot traffic patterns by identifying the areas where there is a consistent flow of traffic between an area of land and a mill and suggesting a “potential link” between the mills and the farms or plantations that supply them in order to get real “ground truth”.

c. The final piece to the solution is the underlying technologies, i.e. ‘Geospatial Data Analytics’ going mainstream. There are three key elements - costs, capability and timing.

The proliferation of affordable drones and satellites is generating unprecedented amounts of geospatial imagery and this increased supply of data is driving costs down. In addition, we have SaaS platforms that are able to pool in demand from multiple end-users and access data suppliers at much cheaper rates. They are able to get wholesale rates given their volumes and constant demand. As you would expect, other cost characteristics include cloud computing and artificial intelligence.

Current practice is simply not scalable and efficient but introducing scalable technologies is improving capabilities. Analytics teams at enterprises and consulting firms normally take months of work and PhD level expertise to derive meaningful insights from crunching vast amounts of spatial data. Normally includes data acquisition/ingestion, data quality, normalization, pattern analyses, visualization. Platforms like GO can automate all of these functions and process information in a fraction of time.

Lastly, in terms of timing, based on our work with some of these companies, we believe these technologies are ready to be scaled up and commercialized across industries. The venture ecosystem understands the massive gap and the opportunity it presents. Crunchbase reports over 150 companies operating in the geospatial segment. There have been 375 funding rounds and over usd 1.5 billion raised in equity financing. Half of these investments have been in the last 12 quarters! Some of the top-funded names include the likes of Orbital Insight, Placer.ai, Descartes Labs, UberMedia.

Also, we have large satellite companies & data providers like Airbus, Planet vying for the same pie. They have been expanding vertically to provide data analytics and this is good validation. Having said that, there is a stronger value proposition for these independent platforms that are able to integrate multiple data sources and data types and this would mark the next generation of innovation.

References:

• SpaceNews: Space and the new ESG business climate

• Geospatial World: Unilever, Orbital Insight pilot leverages location tech to monitor supply chain

• Unilever: How we’re using technology to help end deforestation, Sustainable Palm Oil

• Orbital Insight: In Southeast Asia’s Oil Palm Fields, Orbital Insight Teams with Unilever to Fight Deforestation

• AWS blog: https://aws.amazon.com/blogs/industries/cpg-partner-conversations-geospatial-analytics-and-sustainability-with-orbital-insight/

• HBR: https://hbr.org/2019/08/what-supply-chain-transparency-really-means

• Crunchbase: https://www.crunchbase.com/hub/geospatial-startups/hub_overview_default/top?tab=top_orgs