/ decisionsupport

Do you like movies?

Cinema, like many forms of art, is often a reflection of its time. Since the early part of the 20th century, movies have provided a reflection of the individual struggle against adversity - often due to place or circumstance. Futuristic films, a specific sub-genre of science fiction, take advantage of being set in the future - a realm where infinite possibilities of setting, species and technology exists.

In 2002 I took a fourth year course on geographic information systems, taught by the Dean of the faculty at the time, Dr. Peter Keller. The seminar format reviewed current research on spatial decision support systems - complex theories bridging social and data science with technology and management, that provided a solid foundation on which many of our current web based tools are built, and also introduced simple concepts like complexity limits utility.

The class included a term project. My choice leveraged work I was doing in my spare time - dynamic segmentation of high voltage electrical transmission networks was a dry but entirely relevant project. I probably thought a hard problem would make me look special or important to my peers - much more so than the Dean’s suggested topic which drew crickets from the lecture hall. It seemed nobody was interested in his idea of ‘an historical, film review of the timing between geographic technologies being introduced in futuristic films, and their broad availability in industry’.

Not long later, I’d landed a good job one credit short of my degree. I thought to go back to Dr. Keller, to see if anyone had taken up his project idea, and if not, whether he would be open to a directed studies project for my final credit, that I could do in my spare time after work. I was in luck. We settled on some overall objectives, the sample size of movies to be watched, and deliverables for the project. I had just about 3 months to develop the criteria for my movie list, find and watch 50 movies, record and document examples from each, and produce a report and highlight video for presentation at a national conference.

And so I sought out on the project, taking too long at first to set up my study design and get down to the serious business of watching movies. I soon realized what I’d gotten myself into, and took good advantage of the Pic-a-Flic weekly rental deals from their back catalog, watching films and taking notes nearly every day.

332938_309307405770356_1064890680_o

The films I watched were stratified by age, and the simple selection criteria made use of some popularity ratings on imdb.com. Metropolis and Dr. Strangelove stood out from the older selections, Minority Report and the broad time/technical sophistication covered by the Star Trek franchise from more recent times.

In many cases, geographic technologies were introduced in films in advance of their availability in the marketplace. The human ‘computer’ interactions in Minority Report provide inspiration still coming to life in emerging technologies such as VR/AR and digital twins.

More striking than the specifics of the technology, however, was how they were used as plot devices to help characters navigate the geographic dimensions of the challenges they faced. In older examples, like Dr. Strangelove, a croupier rake tracked a marker showing the location of a bomber, somehow moving in real time across the war room table. There was never time to read a manual. Never time to collect more data. Never an occasion where the technology didn’t provide certainty to act. Characters needed an answer to a question. These were decision support systems.

dr-strangelove-still-580

Today people often talk of the benefits of making data driven decisions. All too often, attempts to make this happen fall short. Invariably, semantics of data collection, storage or management churn through time and resources. Feature and functionality requirements from committee based product design results in overly complex systems, limiting the reach and effectiveness of what is useful data, information, and knowledge, if only it was presented more simply.

Over the years, we’ve come up with three guiding principles in the development of successful decision support systems. Specific decision criteria, or rules of the game define what information needs to be considered. Data availability and characteristics (spatial and temporal resolution among others) influence what information can be produced from the data. Context and relationships matter.

Relax and put your phasers back in their holsters: this is where Foundry Spatial comes in.