Search
Close this search box.

TOR159 — Keeping The Lights On with Noah Klugman and Jay Taneja of Gridwatch

noah klugman gridwatch

Listen Now


The availability of continuous power – that is, the assumption that you can plug in an appliance, or flip a switch without wondering whether or not electrons will flow – is a hallmark of civilization. So much so, that a great deal of what holds up our global economy would not be possible without the assurance that we can keep the lights on (or at least turn them on) 24/7. So, what do you do when the power goes out? Do you call your local utility? Do you investigate your community breaker box? Or perhaps you view it as a unexpected respite from the daily grind and simply relax. In some parts of the world, utilities have moved to a system that utilizes “smart meters.” These devices not only allow the utility to control access to power, but they can also alert the utility when power stops. But smart meters are expensive and may not be a silver bullet, especially in emerging economies. My guests for today’s 159th Terms of Reference Podcast, Noah Klugman and Jay Taneja, recognized the limitations of smart meters and have created a different approach to understanding when the power goes out, called Gridwatch. Gridwatch harnesses data captured by the sensors in mobile phones to recognize when a power failure has occurred and notify the local utility. I think just talking bout how Gridwatch was born and what they’ve achieved over the past few years would be enough for today’s show. But amazingly, I think you’ll find real value in our conversation about the future of electrification and how a small start up with a high-value product, is now currently iterating that product to serve needs that have been identified by their users as even more important then getting the power back on. You can connect with Jay and Noah here:  

IN TOR 159 YOU’LL LEARN ABOUT

  • GridWatch, a phone basic way to track power outages, replacing costly meters
  • The team behind it, from its start as a graduate project and the road so far
  • Real-life applications of Big Data and crowdsouring to actual problems in underserved communities
  • The challenges of pitching social value propositions to utilities companies
  • The subversive element of giving communities a broader understanding of the infrastructure of a grid, knowledge usually reserved for the company
  • How GridWatch realized the value of interaction, to make users be willing to contribute to the “large scale citizen science project”

OUR CONVERSATION FEATURES THE FOLLOWING

Names:

  • Zipline (TOR 115)
  • GridWatch
  • UC Berkeley
  • University of Massachussets
  • IBM, IBM Research
  • Kenya Power
  • Takamoto Biogas (TOR 025)
  • Google
  • Haas-Berkeley, Energy at Haas blog
  • Energy for Sustainable Development Journal
  • World Bank’s EnergyData.info
  • Doing Business Rankings
  • Catherine Wolfram Haas
  • Gretchen Bakke’s The Grid
  • Berkeley’s Development Impact Lab, Community Cellular Networks project

Topics:

  • Energy delivery, reliability, infrastructure, outages and blackouts, monitoring
  • Smart grids
  • Power metering, devices
  • Low-cost, phone-based crowdsourced outage monitoring system
  • Smartphones, sensors
  • Big Data
  • Politics of utilities delivery
  • Social business models, prepaid utilities delivery
  • Privacy

Places:

  • Ann Arbor, MI
  • Berkeley, CA
  • Ghana
  • Nairobi, Kenya
  • Venezuela

EPISODE CRIB NOTES

Boston and Ann Arbor, June 2017   03:03 GridWatch’s first minutes “It’s a research project” The hypothesis: Greater visibility in the distribution level of a power grid can be made inexpensively through mobile crowdsourcing Smartphones can detect power outages and restorations No special and expensive gadgets (outage monitors) needed Easy to deploy which is critical in low income, areas “We’re finding a low rate of adoption” People are deciding to forego tapping into electricity grids in poor areas Hypothesized that its because of the grid’s unreliability   06:02 Power is not reliable, how about Noah’s perseverance? “An interesting problem” There’s an advanced, multi-billion dollar metering industry Too common for engineer graduates to be able to work with expensive hardware Which is simple to deploy and not that costly But they balloon at scale and make infrastructure renewal prohibitive Even in Ann Arbor But less robust utility companies are at constant risk of outages without much that they can do   09:24 Why grids tho The root came from Jay’s advisor at Berkeley A look at the system show a track of iteration of phone as outage detectors First insight: phone’s connectedness to the grid and network When the grid is off (and the phone stops charging), the GW app sends a report The system tracks for simultaneous reports within an area Of course, people can unplug their phone to take it with them But for that, the phone’s accelerometer will tell if the phone is moving The mic can also detect disturbances as little as 60Hz (fundamental grid frequency) “The app can listen whether the grid is working” Nearby Wi-Fi, SSIDs also, among lots of “tricks”   13:12 Big Data you can finally taste Coming up with tangible examples of data gathering and processing was key to get buy-in from utilities companies Noah is just back from Kenya Berkeley graduate, up for IBM electrical engineering research gig New lab in Nairobi, first in Sub Saharan Africa GW started on the labs, then to homes What would it take to bring it to Kenya Power? What kind of work should be expected?   17:00 Transitioning up from cute to industrial “It’s been part of my graduate work” Making the leap with Kenya Power “Way harder than I thought” Noah pitches to KP Shows costs, user case, ROIs Then off to a hacker space Basically the same pitch They made him aware of the user base requirements GW is valuable is lots of people deem it so “The audience you don’t see in the classroom” They’ve realized the need to go to the people right away   20:08 And this is how an electric engineer becomes a gamification specialist in order to survive GW’s first value proposition was worthless KP could tell general pattern about outages, not real-time feedback on households, neighbors KP would learn about them via social media Highly unreliable GW first thought people would just be comfortable with an app on the back-end users would have no relationship with “People don’t call KP that often” They sort of resign, or expect someone else does Users resisted the idea of GW as a ghost app on their phones “Back to the drawing board” Realization: “A user experience would bring users value” Current app has a map, billing features GW as an experience Invites users into a citizen science project   24:45 UX to scale U(x) GW’s problems change with scale The first challenge was growing the base It was only the first “Why would I install it?” “For… the nice feeling of knowing you’re helping KP?” Thinking about how people would interact better with the app helped make it more useful in terms of data gathering Insight: People have worse problems that KP outages GW started to think ways to reward people   27:46 Where are we now? “Because it’s cute, it gets people interested” People and energy companies want to sit with Jay, launch pilots Noah has got cold calls One from Venezuela, which has been rid with blackouts Quantification lowers frustration It does not really need help from the utility company, useful in places of “sanctioned monopoly” Stephen: It gives ordinary citizens insight about a system that provides to them “We’re not being deliberately subversive, but…” Double-edged sword for companies It reveals lots of issues about infrastructure, management Engineer work is hard People alert outages, a team goes, then…? They can realize there’s no outage, or have a hard time pointing the cause, or realize the problem is orders of magnitude larger than thought “People tend to silently struggle” Precedent: Ghana’s reporting of power outages, people are asked to graph reliability Social and political consequences of quantification   33:34 On massive commercials potential for academic research “It depends on who pull us up” There’s lots of excitement, and questions about scale African environments are always unpredictably challenging Impact evaluations, value in outage response processes How to use GW to contribute to better lives   35:54 And on the future of electricity The Smart Grid in Africa will look like no other Currently smart grids allow to track consumption, ebbs and flow Monitoring is based in USD 300 smart meters that give 2-way communication with companies every few… miles? They’re expensive, let alone data management Emerging countries face higher technology costs, but significantly lower wages Meters are not being deployed at the same rate in SSA People buy prepaid meters, but for consumption, no outage reporting Just like 99% of mobile consumption is prepaid How about adapting prepaid meters for outage reporting Better yet: adapting mobile phones for outage reporting (and prepaid energy purchase)   40:13 On prepaid utilities business models Stephen: Gas generation guest Takamoto Prepaid is a good move for utilities’ balance sheets, avoids large credit lines But cognizant prepaid consumers use about 1/4 of power postpaid do On privacy: part of the research questions “If done wrong, it can be invasive” People should be able to turn GW off, or shut down specific sensors (mic) Tone-down phone sensors are still valuable, but affect confidence level of data What are people willing to accept? Comparing data with self-reporting in social media Customers often report outages on social media including their ID and location Code is open source, anyone can look for exploits, data uses “Privacy is huge concern”   46:01 Privacy hounds are often the worst self-disclosers Stephen: Fascinating behavioral phenomenon “Big Brother lives in Menlo Park” Big Brother can help us live better lives But it has to be with customers along for the journey   47:37 Academy ties “My dream is to make GW stable and useful” “Then give it away” GW is geo-agnostic “This is too much for a graduate student team” Servicing millions require scale, focus “Something less ephemeral than a graduate project” Noah’s still focused on scalability on his research “If Google is listening…” No existing intention to monetize it   49:37 Jay & Noah’s picks Energy at Haas blog Academic Journals: Energy for Sustainable Development World Bank’s Energy Data Portal Ease of Doing Business’ Electricity Section Catherine Wolfram (Haas) on Energy Gretchen B~’s The Grid Berkeley’s Village Mobile Base Station, Community-Managed Cellular Networks Drones: Zipline, High Value Applications, Infrastructure trade-offs grid.watch  
 

Please share and participate

If you have any questions you’d like to ask me, Noah or Jay directly, head on over to the Ask Stephen section. Don’t be shy! Every question is important and I answer every single one. And, if you truly enjoyed this episode and want to make sure others know about it, please share it now:
[feather_share show=”facebook, twitter, linkedin, google_plus” hide=”reddit, pinterest, tumblr, mail”]
Also, ratings and reviews on iTunes are very helpful. Please take a moment to leave an honest review for The TOR Podcast!

Love this show? Tell us about why (or why not) below:

Share the Post: