Today I hope to give you the big vision behind the new inter­dis­ci­pli­nary field of com­pu­ta­tion­al sus­tain­abil­i­ty pio­neered at Cornell. Computers and com­pu­ta­tion­al think­ing have rev­o­lu­tion­ized our lives. The smart­phone is the ulti­mate exam­ple of a uni­ver­sal com­put­er. Apps trans­form the phone into dif­fer­ent devices. Unfortunately, the com­pu­ta­tion­al rev­o­lu­tion has done lit­tle for the sus­tain­abil­i­ty of our Earth. Yet, sus­tain­abil­i­ty prob­lems are unique in scale and com­plex­i­ty, often involv­ing sig­nif­i­cant com­pu­ta­tion­al chal­lenges.

Here is an exam­ple of a com­pu­ta­tion­al sus­tain­abil­i­ty approach for bird con­ser­va­tion. eBird is a cit­i­zen sci­ence pro­gram with over 300,000 vol­un­teer bird­ers who have sub­mit­ted over 300,000,000 bird obser­va­tions cor­re­spond­ing to more than 2,500 years of field work. This is excit­ing, but the chal­lenge is that the data are often biased, main­ly from urban areas.

To incen­tivize bird­ers to vis­it under­sam­pled loca­tions, we devel­oped a game, Avicaching, using game the­o­ry. Birders accrue Avicaching points toward a lot­tery for binoc­u­lars and oth­er prizes. The points are assigned to loca­tions based on the bird­ers’ behav­ior, to induce a more uni­form dis­tri­b­u­tion of bird obser­va­tions.

This has been a very suc­cess­ful game, shift­ing bird­ers to under­sam­pled loca­tions. We com­bined the eBird data with envi­ron­men­tal data, lots of data, and using advanced spa­tial and tem­po­ral machine learn­ing mod­els, and high-performance com­put­ing, we are able to relate the envi­ron­men­tal pre­dic­tors with the pat­terns of occur­rence and absence of the species. 

This ani­ma­tion shows the pat­terns of abun­dance of the north­ern pin­tail for dif­fer­ent months of the year, pro­duced by the machine learn­ing mod­els. The mod­els reveal at a fine res­o­lu­tion the habi­tat pref­er­ence of the birds, which allows for nov­el approach­es for bird con­ser­va­tion.

A good exam­ple of high-precision con­ser­va­tion is bird returns, a pro­gram of The Nature Conservancy with the goal of pro­tect­ing migra­to­ry water­birds in California again­st the drought. The eBird mod­els iden­ti­fied the tar­get areas of the bird migra­tion in Sacramento Valley that are pro­vid­ed to The Nature Conservancy. Farmers sub­mit bids to The Nature Conservancy to keep the tar­get rice fields flood­ed in order to provide habi­tat for the birds dur­ing the bird migra­tion.

This big data ana­lyt­ics and market-based approach has gen­er­at­ed over twen­ty thou­sand acres of addi­tion­al habi­tat for water­birds in California. This is a rad­i­cal­ly nov­el way of doing bird con­ser­va­tion. It is only pos­si­ble because we use advanced com­pu­ta­tion­al meth­ods for high-precision con­ser­va­tion.

Computational sus­tain­abil­i­ty aims to devel­op cross-cutting com­pu­ta­tion­al approach­es. Findings can be trans­ferred across domains. This has been a key dri­ving force behind the dra­mat­ic advances in infor­ma­tion and com­pu­ta­tion tech­nol­o­gy, just like the uni­ver­sal com­put­er. For exam­ple, the game the­o­ry mod­el used for Avicaching is also used to decide where to place patrols to pre­vent poach­ing and ille­gal fish­ing. Computationally, the­se prob­lems are sim­i­lar.

We rep­re­sent the­se cross-cutting com­pu­ta­tion­al themes with col­ored sub­way lines. Mechanism design and game the­o­ry is the baby blue line. I talked about the­se cir­cled projects. We have many oth­er com­pu­ta­tion­al sus­tain­abil­i­ty projects cov­er­ing a wide range of appli­ca­tion domains and com­pu­ta­tion­al themes.

Computational sus­tain­abil­i­ty is a tru­ly inter­dis­ci­pli­nary endeav­or since com­pu­ta­tion­al sus­tain­abil­i­ty encom­pass­es bal­anc­ing envi­ron­men­tal, eco­nom­ic, and soci­etal needs for human well-being, for cur­rent and future gen­er­a­tions. A key chal­lenge is how to effec­tive­ly estab­lish the nec­es­sary large-scale inter­dis­ci­pli­nary projects and col­lab­o­ra­tions. Computational sus­tain­abil­i­ty aims to advance com­pu­ta­tion­al meth­ods to help bal­ance eco­nom­ic, envi­ron­men­tal, and soci­etal needs for sus­tain­able devel­op­ment.

Computational sus­tain­abil­i­ty is a two-way street. On one hand it injects com­pu­ta­tion­al think­ing that pro­vides new insights, method­olo­gies, and solu­tions to sus­tain­abil­i­ty prob­lems. On the oth­er hand, it leads to foun­da­tion­al con­tri­bu­tions to com­put­er sci­ence by expos­ing com­put­er sci­en­tists to new chal­lenge prob­lems and new for­malisms, and con­cepts from oth­er dis­ci­plines, lead­ing to cross-cutting com­pu­ta­tion­al prob­lems in com­put­er sci­ence. More impor­tant­ly, it has tremen­dous soci­etal impact. 

Further Reference

Carla Gomes' home page

2016 Annual Meeting of the New Champions at the World Economic Forum site


Help Support Open Transcripts

If you found this useful or interesting, please consider supporting the project monthly at Patreon or once via Square Cash, or even just sharing the link. Thanks.