Alt + E S V

By the Numbers #11

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

Commentary on some of the interesting numbers from ThingMonk Day 2

bythenumbers

(Photo credit: Flickr/morebyless under CC-BY 2.0)

25M

ThingMonk had some awesome talks about IoT and transportation. In addition to the Day 0 talk about automated train interlockings, today Thomas Grassl showed how IoT is playing an integral role in coordinating the traffic of shipping ports.

The port of Hamburg currently ships 9M containers (we’re talking physical shipping containers, my technical friends, not Docker containers) per year, or an average of 24.6K daily. In the next ten years this is expected to grow to 25M containers annually, or 68K containers daily.

thingmonk_tgrassl

Given that the infrastructure surrounding the port (shipping lanes, roads, and bridges) are all space-constrained resources, how can the port possibly accommodate this expansion in throughput? One way is through IoT. (Surprise! I bet you did not see that coming.)

Ports are a complex system with many interconnected parts. The port authority needs to monitor ship traffic, cargo loading/unloading status, truck availability, bridge positions, and potential traffic delays in any of these areas. All of these status reports then need to be communicated and coordinated with the appropriate parties. Using things like geofences and sensors, communications can be pushed to make the system run more smoothly.

1/3

Boris Adryan’s talk made my statistics-loving heart sing.

Boris talked about effective IoT deployments, but his focus was not on the technical but rather on the analytical. Doing the upfront statistical analysis (from surveys to simulations to sampling) can help extend a technological investment.

For example, Boris’ study found that if a collection of parking lots in Westminster had performed a cluster and correlation analysis before investing in an IoT parking solution, they could have deployed their sensors much more efficiently. It turns out that parking lots in the same geographic area display similar traffic pattens and parking space availability was highly correlated with lot size.

thingmonk_badryan

While these findings may seem straightforward, the net result is that with proper upfront analytics, the city could have deployed one-third the number of sensors and still received the same amount of useful information. Analytics can and should play an important role in building IoT solutions.

thingmonk_badryan2

4-10x

Boris’ talk dovetailed nicely with the findings shared by Yodit Stanton. Yodit’s talk was about the role of open data in IoT. In particular, she mentioned that the cost of installing a sensor can be 4-10x higher than the cost of the sensor itself, making it important to strategically use data as a resource. The total cost of ownership of an IoT solution is significant; this supports Boris’ argument about the need for proper analytics and use of data to find the optimal deployment strategy.

Yodit’s company OpenSensors.io sponsored the ThingMonk diversity scholarship this year, enabling 21 members of underrepresented communities to attend the conference. I know I speak for the entire ThingMonk community when I once again express our gratitude to OpenSensors.io for their support of inclusivity.

The scholars contributed so much to the conference. I loved the energy and enthusiasm they added, the questions they asked, and especially their fantastic presentation to close out the conference. Welcome to the community, scholars! We’re excited to keep working with you!

Disclosure: Eclipse, IBM, AWS and RedHat are all RedMonk clients as well as ThingMonk sponsors.

No Comments

Leave a Reply

Your email address will not be published.