Author:
Ms. Chloe Brown
MIT
Co-Authors:
Chloe Brown, Joe Cahaly, Danielle Levin, Eric Wooten, Michelle Kim, Alexis Spinetta

Abstract

Route planning is essential to completing missions in the Army, but it is a tedious, difficult process. By creating a route planning algorithm, similar to Google Maps, we can drastically reduce the effort needed to plan routes. The algorithm will incorporate various factors that are important to operators, such as distance, elevation change, walking routes, and significant activities. Because each person and mission are different, the program will be customizable and multiple routes will be displayed on an intuitive user interface, so the user has different options.

Problem Statement

Currently, when operators are planning the route they are going to take to complete a mission, it is a long and tedious process. The operator first plots the route by hand on a tool like Google Maps or Google Earth. Since they have to consider an environment that is constantly changing they cannot simply put it start and end locations and have the route autofill as it would on a platform like Google Maps. Then, the operator takes the route and translates it to ATAK, which is not user friendly and requires specialized training to use. If given a more efficient program, an operator could use the program as a tool to more efficiently plan a mission, and use the saved time for more important matters.  

Proposal

We propose making a program similar to Google Maps, but with an Army twist. Our program considers things that operators consider when making routes, but conventional route planning tools do not, like elevation change, off-road routes, and significant activities (such as enemy action or threats in a particular area). The routes we generate display on a map, and there will be a route summary including distance and estimate travel time for the operator's convenience. In addition, since customizability is essential to the operators we have talked to, we have a preferences screen where the user can input how important different considerations are to them, estimated vehicle speed or walking pace, and relevant elevation changes. 

We currently have a program that does all of the things I mentioned above, but we have ideas for how we can expand our program in the future. First and foremost, we want to make out program compatible with ATAK, so the operators can use it on their missions. We also want to add the ability to display multiple routes on one screen, so the operator can see multiple options and either pick the one of the displayed routes or have a good place to jump off of. We plan to spend the academic semester working on these problems.

Challenges and Unknowns

We do not know how long it will take to incorporate our program into ATAK; however, other members of MIT ROTC have done this type of thing before, so they will be a valuable resource.

Comments

kcole | 8 February 2021

Love the idea.  Perfect is the enemy of good enough - but some extra bells and whistles might include the viewshed to the points along the path (where friendly/enemy sight-lines occur) and friendly egresses (or indication of areas with difficult egress).  Also, in an envisioned future, one could worry about an adversary using a tool like this to predict likely movements and to set up entrapments, IEDs, or ambushes.

DFreinberg | 9 February 2021

While deployed to Afghanistan as a Commander of a Forward Support Company Commander, my unit could have benefited from this but I offer some comments

How do you factor in route clearance?  There is a system of rating routes based on duration since clearing.  Most often, units use the same routes all the time because that's where the RC units focus their efforts.  Of course you want to minimize predictability but if there's always a gun truck on the road, generally the enemy will have a tougher time emplacing obstacles / IEDs.

Building the convoy in the system you mention would be great to consider turning radii and average speed, potential choke points etc. A convoy with a heavy equipment trailer behaves much differently than one with only guntrucks versus unmarked civilian vehicles.

Not that I'm a huge fan of Blue Force Tracker but that's another area you may want to look.  Rugged system with high Ingress Protection (IP) connectors.  Touchscreen, etc

dBlocher | 9 February 2021

I also love this idea, and agree that making it compatible with (or coding it as a plug-in for) ATAK would be great since it sounds like that will side-step the problem of stove-piped systems. I also agree that offering the user an option or options would be good. With Google maps you are invariably optimizing drive time with or without avoiding tolls but soldiers likely have more complex optimizations where input parameters may exist on a sliding scale such as risk - i.e. if out on a routine and non-time critical mission perhaps you would use recently cleared routes even if the time to destination was long, however if your timely arrival was mission critical you might accept more risk in taking a quick route that was not recently cleared. It sounds like you have a good handle on what the parameters are though I'd suggest thinking about how the would be user would interact with those parameters, and then consider reccomender systems - i.e. code the algorithm to offer a set of options from which the user selects, and possibly alters to their liking.

kgregson | 11 February 2021

Nice idea.  Something that you might also like to take into account for the tool is developing trust and confidence of the operators.  Giving them an opportunity to understand what the recommended route(s) are based on.  If you have access to the ATAK server, you may be able to provide layers they could toggle on/off (like the viewshed and route clearance "heatmaps" mentioned above, recent OPFOR activity, and others) that the recommendation is based upon.

awright | 19 February 2021

Great idea. I'm wondering whether you would want to focus on an easy-ish detection problem (as opposed to the video surveillance example you describe) so that you can focus on the heterogeneous data sources you would need to integrate to serve multiple routing optimization factors.