Point Me To It
Originally published at blog.atherenergy.com
When we decided that we had to have Points within 4 km from any location in the city, we got to work on figuring out how to make it happen.
But before that, if you don’t know what a Point with a capital P is, please watch this.
Realising it’s a Complex Problem
Let’s start with the simplest of approach. Let’s draw a grid made of squares with sides of length 4 km. Placing a Point at every corner of the square should result in no point in the city being more than 4 km ( even lesser if you bring trigonometry into it, but then it won’t be the simplest of approaches anymore ) away from a charging location.
That approach may also fail because the roads themselves aren’t arranged in a grid, for instance, there are one-ways, turns are blocked, and there are u-turns to get to a location on the other side of road. This means we may still end up having locations in the city which are farther than 4 km away from a charging location.
One way to mitigate this would be to reduce the length of the edge of the square from 4 km to a smaller number that has enough buffer to ensure that no point is more than 4 km away. However, there is no way to be sure that even after drawing smaller squares, we meet our 4 km criterion. Furthermore, the smaller the squares in the grid, the more Points we would need to set up, which could subsequently result in some of them becoming redundant and barely getting used at all.
It would not be a smart way to do this.
TL;DR, we had to find a way to keep the number of Points low enough for it to make business sense, thereby allowing us to continue the regular supply of Nutella for the employees, but not go so low that we couldn’t meet our criterion of having one every 4 km.
Solving it Properly
To answer this question, we decided to write an optimisation program. This progam would calculate where we should put Points and how many of them we would need to cover the city. The program took as inputs -
- Bengaluru’s road network with various other details such as which roads are one ways, which roads are not available to motor vehicles, etc.
- The requirement of having a charging station every 4 km,
- Candidate locations for setting up Points.
The program would then go into deep thought and take a fraction of 7.5 million years to come up with an answer which was close enough to 42 for the programmer to be overjoyed. The AtherGrid team was also overjoyed because their estimate for this was about 100 charging locations and they were happy that they now had to set up about 60 locations less. The scenes were heart-warming. Cue the hugs.
The solution also listed out an optimal combination of candidate locations, about 40 in total as mentioned before. That is to say, the program had calculated that if we were to install Points at each of those locations, we would satisfy the criterion of one being able to reach a Point within 4 km from anywhere in the city.
Our AtherGrid team then used this information to seek out potential partners and check whether their establishment met our other requirements, such as having parking space and the ability to handle the extra electric load. If the site did meet all our requirements, our team would initiate discussions with them about setting up a Point there.
Based on how these discussions went, we had to keep updating the optimisation results. For that, we had to add two more inputs to the program:
- Candidate locations which were ineligible, or had declined our offer of setting up Points,
- Candidate locations where we had agreements to install Points.
The exercise took about 2 years as the team updated these inputs, talked to partners, and iterated.
The maps below show the final AtherGrid that is now being set up. Red marks areas farther than 4 km from a Point, and green marks areas with Points within 4 km. The programmer is not a designer so pardon the aesthetic and the aspect ratio.
As of today, we plan to set up the 33 charging locations ( and 60 Points ) marked in this map. With these 33 stations, you will be able to reach a Point within 4 km from about 85% of the city’s roads. We are continuing our efforts towards ensuring maximum coverage in Bengaluru and will be replicating this approach in other cities that we are planning to launch in in the near future.
