Solar for the home

Disclaimer: I am no expert in Solar power generation, nor am I an Electrical Engineer. So what I am about to request/propose might be the most absurdest thing – but for me, ignorance is bliss, and I shall propose.

The biggest problem that I see about solar power is that it is available only during the day and that is not the time we consume most of the power at home (work-places is a different issue – I am addressing only homes here). We need a lot more power at night to light up our homes, drive our fans and ACs, and our televisions, and routers and what not. And for doing this, we need to store this energy in batteries.

A premise that I make here is that a lot of us have inverters and they have batteries in them. But with this inverter arrangement, while we keep our batteries fully charged, we use them only if we lose power. Else it is lying unused (dormant).

1. Decouple the battery from the inverter.

2. Split wiring at home into low load and high load. For those who are using an inverter, this would have already been done by now. Keep the high load circuit on direct power (from the utility company).

3. Charge the battery during the day using solar panels. Once the battery is fully charged, divert solar power to the low load circuit to run any day time uses such as fans, TVs etc.

4. For night time power in the low load circuit, use the power stored in the battery. If the battery goes below a certain threshold, switch to main line from utility power company. If there is still power in the battery, even after day breaks the next day, continue using power from battery, switch to utility company only after battery discharges

Couple of advantages that I see here:

1. We use the power that we generate in the day time first, and then, only if needed switch to main line power in the night. So there is definitely a big component of night time power consumption that is offset. If load is low, we can perhaps offset some day time power consumption too.

2. A battery that is charged and discharged continuously lasts longer.

While a lot of this is based on my imagination, I am pretty sure this is doable in the realm of current age electrical circuitry. 

My question: Has someone done this already? Any commercial solutions? 


One IoT app that I want right now ..

After I wrote the Google I/O post that I wrote today, I got ready and drove down to work. And while I was driving down, I was thinking about one immediate IoT application that I want

Disclaimer: This may be specific to Indian users.

trolley For those who have not seen this picture on the left, this is a Gas Cylinder trolley with an integrated weighing scale. Several brands market them now (including Pigeon, from where I took the picture from). It serves two purposes. Check if the cylinder is full, when delivered – there was a time when there were some agencies scamming the consumer by delivering half full cylinders). The second purpose was to know when the cylinder was close to getting over – before it actually got over (the weight of an empty cylinder canister is published information). 

My dream IoT app would connect this trolley to the online refill booking system of my gas cylinder supplier. Typically it takes time for the distributor to deliver the cylinder. The app would continually track the usage of the cylinder and the wait time of the distributor and order appropriately such that I get the new cylinder close to the day when my cylinder gets over.



Google and IoT – Brillo and Weave


The thing that caught my attention in yesterdays Google I/O conference was the new (renewed?) focus on IoT. And this time, it looked like Google went back to basics. Back to the drawing boards. I like this idea. The Internet of Things (IoT) is not something that is going to be a fad and move into oblivion or mainstream in a few years and forgotten about. It is something that is going to grow over time. There will be an ecosystem around it. It will get smarter. It will become more pervasive. And it is going to stay for a long time.

The Internet of Things (IoT) is not something that is going to be a fad and move into oblivion or mainstream in a few years and forgotten about. It is something that is going to grow over time. There will be an ecosystem around it. It will get smarter. It will become more pervasive. And it is going to stay for a long time. The applications are many. From smarter homes to smarter public transportation systems. From cleaner cities to more energy optimization. From personal convenience (sometimes bordering on creepy?) to mainstream production. From retail to manufacturing. Everywhere. And here is where I get to make an almost sci-fi’ish statement – “Here is where machines will talk to each other.”

For this to happen, the basics need to be right. Almost like how Android started.

Google has introduced Project Brillo, which it claims to be the operating system for the IoT. It is a low-level layer of Android which can be ’embedded’ into the machines that are going to be part of the eco-system. These could be your Nest thermostat, smart LED bulbs, or in the future – smart ovens, smart refrigerators, smart dead-bolts (door locks) etc.

The second part of the puzzle is a standard communication protocol by which these machines will talk to each other. Google introduced a new protocol called “Weave”. (Somehow these names also seem refreshing opposed to IEEE 455.34 or something similar). The weave protocol will be a standardized protocol, which will enable any machine to talk to another machine, or to a phone, or to the internet — which is pretty  much the internet of things. This where the magic happens. This is how your phone and your car will tell your coffee maker to start brewing and tell your thermostat to warm your living room, when you are 5 minutes from reaching your home.

Google is also going to be starting a Weave certification, so that third party product manufacturers can make weave ready products. This opens up the ecosystem that I was talking about. Guys like WeMo and Belkin will jump on to the bandwagon first. And then more common-place companies like the Black and Deckers. And then in parallel, the Fords and GMs.

I am liking this world now.

image src: arstechnica

Zomato and their way forward

Been wanting to write about this for a while. Today seems to be the blogging day. Not feeling sleepy, and feeling an awesome urge to write, and write, and …

As you probably are aware, as of now, Zomato is a restaurant discovery service. And a damn good one at that. They have taken this and scaled it like no one else’s business internationally. Buying UrbanSpoon in the US completed this strategy. They have the US market covered too.

So, now that, people can figure out where to go eat, or order delivery from, what are two things that they could move forward on.

Home Delivery for food. When you can show them where to order delivery from, why not deliver the food itself. This is what Zomato has been experimenting on, for a while (or so I read somewhere). There are already quite a few players in this industry, like the tiny TinyOwl which got some not-so-tiny-amount of funding, and others too (such as delyvery, swiggy, grofers etc). The modus operandi in most of these services is:to display the menu of the restaurant on their website

  1. Display the menu of the restaurant on their website
  2. Take the order from the customer (app or website or phone)
  3. Call the restaurant and create the order
  4. Send the delivery person to the restaurant to pick up the order
  5. Delivery person delivers the food.

The one hiccup that could happen (and happens a lot) in this strategy is the fact that restaurants can say that they cannot make a certain order item, because of chef-absent, raw-material-over, whatever-nonsensical-yet-plausible-reason. In this case, the middle delivery guy is hosed. He needs to call the delivery company, who then calls up the customer, and asks for an alternative order, and pass it back to delivery guy, to the restaurant, etc, and so on and so forth. You get the drift.

Zomato, from what I read, is planning to solve this problem in a pretty nifty way, by using technology. They are planning to give all participating restaurants a Point of Sale (POS) device – most likely an android tablet-like device. When an order comes into the system from a customer (app/web), it goes directly to the restaurant, which has the control to accept or reject order-items immediately. The one inefficiency, that was the delivery-call-center is gone. Once the order is accepted by the restaurant, the delivery guy is dispatched to the restaurant, who picks up the food and delivers. To me, this sounds like an awesome use of technology – well worth the cost of a cheap tablet – many many times over.

Now, let us step back. There are several more side-advantages of this approach. Zomato now has a foot (well, a technological foot) in the restaurant’s door.

  • Zomato and the restaurant can capture analytics on which food is most ordered.
    • Zomato could potentially even crunch the numbers and figure out how long a certain delivery might take, based on distance of customer from restaurant, timing of the order etc.
  • Zomato can run online specials in collaboration with the restaurant.
  • Zomato can feedback ratings from customers back to restaurants directly.
  • Zomato ratings for restaurants and the ‘price for two’ numbers can become accurate.

Table Booking: Now lets talk about the second direction in which Zomato could potentially expand. If you are a restaurant discovery service, and you have suggested a restaurant to a customer, who has narrowed into this choice, based on a variety of choices, the next easy thing to capture is to help the customer to book a table.

Sure, there are players in this market too –, etc.

What does Zomato have (or can do) that gives it an edge?

Well, let us go back to the first direction (home delivery) that we discussed in length earlier. We remember that Zomato was giving the restaurants a POS (point of sale) terminal – an android device, which was going to help in getting an order into the restaurant. This same device could very well be used to book a table in the restaurant. In fact, this booking would be so much more reliable and real-time. This would replace the register that the Maitre’D has on his podium at the entrance. So, when half the restaurant has been booked by a large party online,  someone showing up at the door, would know immediately ; and vice versa.

With such a real time interaction between the restaurant and Zomato, one could do several more things as well:

  • Accurately enter the number of regular chair seats vs sofa seating and track the reservations based on this. (Personal problem that I have faced often).
  • Crunch analytics on when a restaurant is most crowded, how long (on an average) a table is occupied, and hence when a next slot can be opened up.

And ofcourse, let us throw in some capitalism for the fun of it:

  • Zomato could have ‘promoted’ hotels –
  • Restaurants could do dip pricing (my invented term – opposite of surge pricing ) for irregular hours
  • When a customer tries booking a table, and is unsuccessful, Zomato could suggest something else nearby. (This could also be the case, when you have booked a table, but the restaurant is so busy, that they cannot guarantee you, and make you wait).

Summary: I have been a big fan of Zomato and what they have been doing. And I see these two directions as promising directions that they are heading (could head towards).

Well, I have to say this. I do have one thing I did not like that Zomato did. This was when they moved from Bangalore to Gurgaon. They tried trash talking Bangalore to lure devs to Gurgaon. While it got them a lot of publicity, I kind-of felt that was in bad taste. Well, to each, their opinion.

IoE and Retail

I read an article by Mala Anand (@malaAnandCisco) of Cisco on how Internet of Everything (IoE) can be harnessed for retail, and that got me thinking. This is another of those dream product posts that I occasionally write (dream-write?). There may be overlaps in some of the things that Mala had used, but this here post is my interpretation 🙂

My definition of IoE or IoT is the system where multiple devices capable of collecting, transmitting, and/or processing data, are interconnected towards achieving a specific result.

Let us take retail as an example, for the case of this blog post.

As you walk into the door, apps from your mobile phone could communicate to the store about the list of things that you are looking for. Perhaps at a high level, you walk into a super-store such as Star Bazaar in India, or Walmart in the US, and your intention is to buy some groceries and perhaps a dress for your nephew. One of these things could be done:

  • Your to-buy list could be transmitted to a central unit, which could dispatch a human personnel to you to guide you through buying your dress.
  • The store central unit sends the directions to your app indicating where the items you intending purchasing are stocked. It is worth mentioning here, that stores tend to keep changing their locations. So without the latest directions, you could be searching for a while.
  • If you have a long list of grocery items, the central system could guide the phone in your app to an optimized route.
  • If you have taken a shopping cart/trolley, the central system could now talk to a display on the trolley and guide you.
  • And ofcourse, let us toss in a bit of capitalism in here, judging you by your past behaviors, the system could ‘surreptitiously’ guide you to some items that you did not have on our list, but you crave/wish for. And ofcourse give you some deals while at it, to persuade you to buy it.
  • The system could also help you keep you within a budget – money-wise and time-wise.

And hey, while you were still at home, creating your to-buy list, your refrigerator could have let you know the status of milk and other staples at home, and if you need to purchase any.

While at the store, let us shift away from the customer experience, and move towards how it could potentially help the store itself.

  • Based on the sensors that could detect weather (current and predictions), the store manager could get data on if he should be stocking up on related items such as umbrellas, suntan lotion etc, and if he should be putting them up in the center aisles.
  • The temperature sensors could also optimally adjust the thermostats and lighting (this is already done by NEST kind of devices)
  • Stock inventory (weight sensors on the shelves) and Resource deployment could also be updated based on connected devices in the store.
  • Billing lines could be avoided – the connected devices would have calculated the bill, gotten the buyers approval and charged the credit card or deducted from an online wallet.
  • Connected cleaning agents such as robotized roombas could help clean the stores in co-ordination with the other systems in the store.
  • Connected cameras could potentially infer demographics of the customers visiting the stores. Example – Retired folks in a Target at St.Petersburg, Tampa would probably not be needing baseball bats and specialized soccer equipment. Golf clubs might be a better idea to stock.

These are some of the ideas that came up, off the top of my mind. But the Internet of Things is a very interesting subject and even though, much of it is made of dreams, there are promising directions that companies are headed towards, which makes all of these dreams potentially feasible in the not so distant future.


Facebook New stories : Mobile to Web

As I have said earlier, I notice UI/UX changes. Some earlier picks herehere and here. I just noticed something new today in Facebook. I have seen this in their mobile app, but one of the first times, I am seeing a company move a good UI feature from an app to the web.



I have not seen the “New Stories” button in the webpage before. New posts would either auto-load, or I would need to go click on the f button to load new pages. This brings in a new cognitive feature to show that there are new stories.

Good stuff, Facebook. I love the way you are moving features seamlessly between web pages to app and vice versa.

Twitter Trends – UI Change?

A little while ago, I wrote a post on the awesome new feature in twitter web UI, the quote-and-RT. I am loving it But before the euphoria died down, I just noticed that, twitter has made yet another UI (or maybe trying out on a subset population). When I log in to twitter today morning, I see a dense trends column on the left side of my screen.

2015-04-23 10_17_24-(8) Twitter

To be honest, I do not remember where the trends column used to be before – I rarely look at what is trending. But today seemed to be different. I noticed it because it is crowded and ugly. It is crowded because of the extra text below the actual hashtag. It distracts me from my timeline.

The fact that I never used to look at the trends (and hence did not let it affect/influence) my conversations on the timeline, makes it even harder for me now.

There is definitely a cognitive distraction that is happening here and I am not liking it. Are any of my readers seeing this as well? Are you OK with this?

And deep inside me, there is one more fear in me. Was this a means to drive people towards trending phrases – which could potentially be sponsored. If so, this is nothing short of discrete native advertising. Not that I am saying it is wrong to do advertising (twitter is a publicly listed company after all, and it has to make money), but hopefully not at the cost of distracted users. The last time something like that happened to me was when there used to be these blinking bling pop-up ads on webpages.

Now, has someone written an extension to hide the trends column?

Update: Ok, I found a way to get it back to normal. Click on the ‘Change’  next to the trends. It will ask you for which city you want trends for. Type in your location. There is a button called “Tailored Trends”. Do not click on that button, and just click Done. It will go back to a shorter trends column with just hashtags. Phew.

Ecommerce kiosks

I had toyed with this idea earlier when I wrote the “Ideas for Bigbasket” post. The idea then was to have these internet enabled kiosks where folks could order stuff from big basket and perhaps even collect stuff from these kiosks/counters. Fundamental premise at that time was this was still for the yuppie crowd, but for those who do not want to sit on a computer and order, and most definitely do not want to wait at home for an order to arrive.

More recently, on reading this article about Flipkart going fully the app way (and most likely closing the route), this idea popped back into my mind. Kiosks as I mention above are probably best served using an android interface, and what better news than an ecommerce provider going fully the app route. (Sure you could do this even if it was not fully the app-route, but this was what caught my attention!).

So, let me try and use my static brain a little here, this fine tuesday morning, to see if I can come with a feature set for a kiosk like this.

1. Kiosks – An ATM like machine with only the ecommerce app loaded on to the screen, where customers can order. There could be one or many depending on the traffic the place sees. This would have to be a secure room such as those found for ATMs. Solid power backup and reliable internet connection would have to be a must. The kiosk must have the capability to be open 24 hours. This gives access for someone to pickup their order even late at night.

2. Positioning of kiosk My thinking has changed since my Bigbasket post. I think this kiosk can be anywhere and everywhere. It just needs to be high traffic. In a place like Bangalore, it could be in the BDA complexes. It could be in the MTC bus depots (like the T.Nagar Depot) in Chennai. It could be in the minimarkets (like Brahmaputra market – Sec 29) in Noida. It could be the megamarkets (such as Sarojini nagar Market) in Delhi. Well, these are the places I have lived for significant periods of time, but you get my drift. This is no longer for the yuppie crowd.

3. Manned vs unmanned kiosks: There could be manned kiosks where registrations can be facilitated by a personnel. These manned kiosks could potentially also take cash payments. Unmanned kiosks would have no personnel and could only be used order and take delivery of items using debit/credit cards. Unmanned kiosks could be 24 hours but manner kiosks could be open for a limited time. No special training for personnel would be required, since a current delivery logistics person could very well be reused to man this job- since the job function is exactly the same – except he does not go door to door, but deliver at a single point.

4. Access and payments would have to be through a card facilitated through the ecommerce company and potentially a bank. These cards could be applied for either online or through a help person sitting in the kiosks. These cards could be either preloaded debit cards or credit cards. Bank faciliated cards would take off a liability from the ecommerce company and help greatly on processing payments. Top up on cards can potentially be made easier with top-up cards similar to the pre-paid phone cards that are very popular now.

5. Pickup The access card could be used to unlock a locker, where orders can be picked up. Perishables have a timeline before which one needs to pick the order up. SMS alerts could be set up for locker opening and closing and when the order is available in the locker for pickup. This locker concept has already been tried out in the US by Amazon and is technically feasible. I do not know how successful the experiment was, but cultural significance matter and we cannot rule out it not working here.

6. Ordering would be easy. The card would authenticate them. They browse and choose the product they want to order. If the pickup location is a manned kiosk, COD would be a possibility, else, it would have to be credit/debit card payments.  The existing UX for the app could very well be reused. The delivery location however could have a default location, which is the kiosk location itself. The customer can choose to order here and deliver elsewhere as well.

Advantages to the ecommerce company:

1. Logistics is simplified to a great extent. It can be potentially much faster, since there just needs to be one trip from the motherhub of a city to the kiosk location. Cost reduces because of this simplification as well.

2. Reach is significantly expanded. People who do not have a stable internet connection at home, or a smart phone with an internet connection can easily order stuff off the web now.

Disadvantages to the ecommerce company:

1. There is one more entity that needs to be managed now – the kiosks.

2. There is always a risk of vandalism and pilferage. This needs to be managed carefully.

Advantages to consumer:

1. Biggest advantage to the user would be in the delivery segment. He does not need to be at home to receive the package. He can go to the kiosk on his way home and pick it up on the go. The yuppie consumer would probably still be using this phone or web to order, but would probably use this delivery mechanism for ease of pickup.

2. For a not-so-yuppie consumer, this could mean, he could order something off the web without using up his bandwidth on his phone or broadband. He could also do this at a location which he frequents daily (such as a bus stop).

3. For the masses, the reach would be huge. Now a much larger section of society would have universal access to anything they want to buy. Until now, the restriction for a large section of our society is to buy whatever is available in the store that is closest to where they live. If it is not available there, then it is not available at all. This could change all that.

Disadvantages to consumer:

1. The only disadvantages that I could probably think off here is adding yet another card for access to this kiosk, and potentially a long line that is waiting for pickup. But these could be easily worked around by introducing swift and efficient workflows.



On adapting successful UX methods

I notice User Experience (UX) differences and how they affect my productivity. I love products who focus on great UX. I love products who continually evolve their UX to become better and better. You know what I love even more – products who recognize good UX behaviors and adapt it to their own. And I recently came across a fine example of the latter – Twitter.


This is the android twitter app. Do you see the “New Tweets” button at the top. This is very new. Facebook has had this for ages (it is called “New Stories” and it the button has a more oval structure to it). Clicking on the “New Tweets” button lets you know that there are new tweets and that you can click on that to scroll up to the latest tweets. This also saves you a pull down gesture, which is kind-of hard to do if you are holding and operating your phone with one hand (which is a pretty common use case).

My principal point here is that, if you recognize a good UX mechanism, it is my personal believe that, there is nothing wrong in adapting the mechanism to your product (unless it is patented ofcourse). It helps standardize UX across classes of apps. There is also a sense of sharing between the companies. I am sure FB spent quite a bit of UX effort coming up with their equivalent.

Request: As always, I have one request, which I am sure Twitter will not see, but that is fine, I will indulge myself. I would love to see the “New Tweets” button enhanced with the number of new tweets  – example – “132 New Tweets”. Twitter has the underlying algorithms for this, since it is present in their webapp.