The American Meteorology Society is celebrating their 100th annual meeting this week in Boston with a 4 day conference featuring leading minds from the industry.
While the conference features a number of presentations, workshops, and panels, one of the most compelling sessions took place at a town hall discussion Tuesday afternoon on how the cloud can be leveraged for scalable operational meteorology and what the applications could be for businesses, academia, and the general community alike. Tomorrow.io’s Chief Scientist, Daniel Rothenberg led the discussion along with esteemed experts from Google, Amazon Web Services, Microsoft Azure, and the National Oceanic and Atmospheric Administration.
Panel of Featured Experts:
- Daniel Rothenberg, Chief Scientist at Tomorrow.io
- Tim Carroll, Director, HPC & AI for Research at Microsoft Azure
- Shane Glass, Program Manager and Lead for Google Cloud’s Public Dataset Program
- Kevin Jorissen, Lead for HPC and Research Computing at AWS
- Irene Parker, Chief Information Officer for NOAA Satellites
- “Society is asking the weather community to step up its game.” – Kevin Jorissen, AWS
- “Forecasts aren’t inherently valuable. They become valuable when they inform decision-making.” – Shane Glass, Google
Throughout the town hall, a few core themes continually surfaced from both the panelists and the audience given the evolution of the cloud, including:
- Artificial intelligence
- Machine learning
- The role of human vs. machine
- The state of weather data
- Commercial applications
Let’s dig in:
- Focus on action, not the forecast: The state of weather data was highlighted by most panelists, and the overwhelming response was that we’re at a point in time where we have the data we need, and can now allow people to focus on problem solving. Additionally, we’re spending too much time focused on technology and not enough time focused on the science and solving problems. People and businesses need to get past simply getting the forecast and focus the majority of efforts on what the forecast means, the impact, and how to convey those insights to inform better decision making for both people and businesses.
- The cloud opens the door to everything: The impact of cloud capabilities vs. on-prem is truly a revolution and one of if not the most exciting advancements happening right now. With the cloud comes access, which creates adoption, collaboration and innovation. Furthermore, computing and operational limitations (including financial) of on-prem can be a thing of the past given the elasticity of the cloud.
- Weather for x: The days of “Uber for X” don’t feel so long ago, but similar sentiments were shared that “weather for x” is going to be a significant opportunity moving forward. For instance, measuring how the weather impacts human health conditions and being able to draw links both in real time and historical. Advancements like these are ready to impact society in significant ways as weather continues to play a bigger role in our everyday lives. We’ve never been able to think about weather in this way, but revolutionary software offerings (insert plug for Tomorrow.io here) allow us to expand our vision of what is and what can be.
- The impact of artificial intelligence and machine learning: With better weather insights comes bigger demands from consumers, governments, and businesses. People want better weather data, faster, and always in real time. They want historical information, current analysis, and future predictions they can rely on. All of that sounds great, but how do we do it? Scientists are people too, and we can’t just replicate them at scale…and that’s where AI and ML come into play. There’s no doubt AI/ML will play a major role in the coming years, but people shouldn’t worry about humans being taken out of the equation. Humans will always need to be in the loop and are actually a critical component to be able to provide feedback to models. A helpful metaphor given was to think about how NASA is trying to use AI/ML to automate a bunch of tasks astronauts have to do during flight, allowing them to focus more on the science. The same concept rings true for weather, if our AI/ML can automate redundancies and allow our experts and businesses to focus on science and operational problems that need to be solved it will be a more efficient use of our time by an order of magnitude.
- Start with your problem, nothing else: The final point here is that given the current advancements of weather technology in addition to the cloud, businesses need to approach strategy in a different way. You don’t need to focus on what you think you need built specifically or what your budget is, just start with the core problem or problems impacting your business. Expand from there to include all the details, move to why fixing this problem (or problems) is critical to your business and what the impact would be, and finally consider what the impact would be operationally. This is a completely different way to have a business conversation than in the past and is much, much more helpful for all parties involved.
Thanks to AMS for putting on such a great event and to all of our panel guests for joining. If you’re in Boston don’t forget to check out the other great sessions over the next few days listed in the conference agenda. Cheers!