I've just submitted a grant proposal on "Functional multi-objective optimization under uncertainty" to the Swedish Science Council. This is in the same general direction as the proposals submitted (and rejected) last year (Exploring fair climate policies under uncertainty and conflicting goals: a computer-aided approach, Computing optimal policies for clean energy and sustainable development, ...) but I'm hoping for better luck this time.
The proposal fits well with growing community around "Programming for the Climate (PROPL) (see the blog entry about our talk "Can computer science help climate policy making?") and the ongoing PhD course on "Functional Programming and Climate Impact Research".
Title: Functional multi-objective optimization under uncertainty
Abstract:
The latest IPCC report clearly shows that avoiding unmanageable consequences of climate change requires a profound transformation of our society. But the report provides little clues on which concrete climate policies can promote and sustain such a transformation.
This is not very surprising: designing fair carbon taxes, and effective schemes for trading emission rights is difficult. And bans on cheap, emissions intensive technologies are difficult to agree upon, and hard to implement.
In a nutshell: a transformation towards a fossil fuel free society must itself be sustainable. It must reach the national climate goals but also ensure safety and prosperity of future generations.
This requires an open, iterative decision process that integrates climate science with empirical data, domain-expert knowledge, etc. into stylized decision problems.
These problems need to handle uncertain knowledge. But they also have to account for uncertainties in the decision process itself: decisions are often delayed or postponed.
Doing so requires a method that can deal with different kinds of uncertainty rigorously. The project will develop this methodology by focusing on how to ...
1) encode stylized decision processes?
2) encode preferences and goals?
3) explore alternative policies?
We do this through specification, prototype implementation, and verification of multi-objective control problems under uncertainty, and by building domain-specific languages for exploring their solutions.
Research plan: Jansson_VR_2024.pdf
Publication list: Jansson_publications_2024-04.pdf
The proposal fits well with growing community around "Programming for the Climate (PROPL) (see the blog entry about our talk "Can computer science help climate policy making?") and the ongoing PhD course on "Functional Programming and Climate Impact Research".
Title: Functional multi-objective optimization under uncertainty
Abstract:
The latest IPCC report clearly shows that avoiding unmanageable consequences of climate change requires a profound transformation of our society. But the report provides little clues on which concrete climate policies can promote and sustain such a transformation.
This is not very surprising: designing fair carbon taxes, and effective schemes for trading emission rights is difficult. And bans on cheap, emissions intensive technologies are difficult to agree upon, and hard to implement.
In a nutshell: a transformation towards a fossil fuel free society must itself be sustainable. It must reach the national climate goals but also ensure safety and prosperity of future generations.
This requires an open, iterative decision process that integrates climate science with empirical data, domain-expert knowledge, etc. into stylized decision problems.
These problems need to handle uncertain knowledge. But they also have to account for uncertainties in the decision process itself: decisions are often delayed or postponed.
Doing so requires a method that can deal with different kinds of uncertainty rigorously. The project will develop this methodology by focusing on how to ...
1) encode stylized decision processes?
2) encode preferences and goals?
3) explore alternative policies?
We do this through specification, prototype implementation, and verification of multi-objective control problems under uncertainty, and by building domain-specific languages for exploring their solutions.
Research plan: Jansson_VR_2024.pdf
Publication list: Jansson_publications_2024-04.pdf
