Dispatch enables objects and goods to navigate on their own through the world, without humans at the helm.

This is one of many projects I worked on at IDEO CoLab.


Dispatch is an autonomous box that gives identity and agency to the contents it’s carrying. Dispatch enables objects and goods to navigate on their own through the world, without humans at the helm. In a Dispatch system, a crate of strawberries could route itself to a nearer store if it identified a risk of spoilage, a crate of precious metals could dynamically elect for a less risky route, or a fragile piece of equipment could pay out for delivery at a rate directly proportional to the quality of its treatment during transit.


Dispatch was part of a 7 week project that I led with two other IDEO fellows. Our main prompt was to design a venture that was build off of a DAO or Decentralized Autonomous Organization. This brief was quickly narrowed once we began to iterate venture ideas. We choose the use case of transportation as our design challenge because we felt it would be a good lense to explore products and would be quick to prototype in. The main outcome of this project was to better understand the commercial value of a DAO.

First Iteration

We went through a number of tests many of which were explorations in our core assumption, can we pay people to move packages for others when it’s not clear what’s inside. We went so far as to hire people off of craigslist in order to test this assumption, to our surprise it worked. Still we wanted the boxes to move themselves so we snuck some handmade boxes with bounties and a map onto bikes at a bike share. This was a direct test if we could give a box the tools it needed to navigate on it’s own. Again to our amazement some of the boxes made it home.

Findings and Changes

We were quickly able to validate the fact that the boxes could pay people to move them, so we decided to move on to the next series of assumptions we were making. What can the box do to mitigate theft or damage and how does it keep itself safe?

Second Iteration

Our answer to this was to create a box brain so to speak that could track simple parameters, is the box to hot? Is it being shaken? Is it being taken to the wrong part of town? With it monitoring it’s own state, the box would rank it’s courier, much in the same way you rank an uber driver. The catch is that it would do this in real time so as the courier you would know that the box does not like the way it’s being treated, and you’d know how to change that.

Final Iteration

Of course this is a challenge to showcase as a whole system so we designed a simple interactive installation to communicate our work to the client. In the installation you pick a box with a screen and move it from shelf A across the room to shelf B. Once on the other side the box pays you. A simple but effective way to communicate the core of the idea. If you started to dilly-dally the box would protest and threaten to dock your pay.

Venture Viability

Unfortunately but not surprisingly it’s not terribly cost effective to put a micro controller in every shipped box on earth. Still the core functionality and the principles remain the same. Using distributed networks like the DAO and the internet of things it is possible to build systems that can monitor themselves and automate payout based on how well a job is done.

If you have feedback please feel free to email me by clicking below. I'd love to hear your thoughts.