There’s already a great start in terms of online material describing the new VMLCI initiative:
I wanted to open the door slightly further with a few thoughts from engineering and kick off a series of blog posts describing some of our projects, progress, and thinking.
Our foundation starts with an unusual diversity of talent, projects, and opportunities.
I’m not aware of many organizations that attempt to take on a pipeline of open, unsolved challenges spanning such a diverse range of technologies and domains as we do. A fair number of our projects and initiatives continue to involve machine learning (ML) in some form.
Take a moment and think of the broader set of solutions creating actionable intelligence when it comes to:
- large geographic areas of concern (land, sea, etc.)
- bad actors (poachers, illegal fisherman, etc.)
- various, dynamic situations (disasters, pandemics, etc.)
With limited resources, time, and a plethora of sensor data. This is the equivalent of bringing military style intelligence and insights to non-military priorities.
Solving for these challenges involves developing a vertical stack and horizontal suite of sensor (EO, IR, RF, acoustic; SAR, etc.) technology - from satellite constellations to one of the most robust, non-military BVLOS drones out there (developed in-house) to various terrestrial deployments.
These are merely a tiny snapshot in time of use cases involving ML at Vulcan. Apples and oranges, but worth emphasizing all of this comes with Vulcan’s ongoing executive support, funding, ambitious scope and endurance not found with most venture-capital-backed startups and conventional, for-profit technology companies.
This isn’t business as usual.
For those of us who come from a more traditional product, business, and for-profit background, the relevance of conventional metrics (market share, revenue or other elements of valuation, etc.) can be an ongoing challenge even when we’re not attempting to use them as a proxy for philanthropic, conservation, or related impact.
It’s important not to fall into the hero mentality trap of past successes. Applying the same rigor and methodology that achieved results in one domain may not work at all in another domain. But it is still critical to challenge the status quo, no matter how long it has been in place. Respectfully.
Regardless of this particular challenge, there’s a fundamental draw with most engineers when it comes to seeing any of our solutions gain genuine traction and impact at scale and with endurance. At any given point in time, we ultimately know when it’s working versus when it’s not - though second (velocity, in all its forms) and third derivatives can be a bit more tricky.
And then there’s taking “sustainability” to the next level. To us, this doesn’t involve just breaking even when it comes to operational expenses! This one deserves its own, future blog post.
And even though we’re just getting started….
There’s already remarkable work and global impact, which speaks for itself via past Vulcan projects and the various Allen Institutes. The same ambition and support is found behind the genesis of this latest effort.
We’ve already dipped our toes into this #ML4Good work with a number of solutions, which already are or will soon be making a difference in the real world. Again, moving forward, we’ll post more around specific projects, areas of interest, and so on.
In the meantime, even from those who are managing their “grass-is-greener” impulses or feel the next AI Winter can’t arrive soon enough - please reach out and Contact Us! There are so many ways to contribute, even beyond in-house, full-time engineering positions - collaboration through partnerships, consultation, open-source style projects, and more early stage brainstorming and ideation events.