Physical Computing @abhiarduino - Tumblr Blog

Posts

A simple video of my project working. I have a few more modifications to do, but I am only able to upload my video on the NYU campus (Internet is kind of slow) so I'll upload this video now, and then upload a few pictures of a few UI changes I do later.

Documentation a little better:

Project is called District. You go into a particular district and you're able to see places to eat within that district, which are the best or the closest. It is a way to explore a particular district without actually having to pull out your phone, and planning your time in that district easily. If you're a tourist it will give you a simple way to understand where you are and be able to make plans accordingly.

This will basically help you plan your time within this area of the city, and help you explore it in more depth. I have only accomplished the food section because this was a really hard problem to solve for me, and I think I have done well with it. The problem with Food and restaurants is that there aren't many APIs that you can call or use to be able to get details about restaurants and query them easily. So I had to implement my own scrapers and scrape from websites like Open Table, Seamless, Delivery.com, Grubhub, etc. This was a complex problem as I had to write algorithms to get then parse data within these sites, and then serve it for the user to use very fast.

My UI is basically a kind of a joystick with a Magnetic Stripe reader (For now it's a RFID because I only had that). The user will swipe their Metro Card or RFID card and then gain access to their profile (it will remember you based on that number). This is already implemented for the RFID but I wasn't able to secure a Magnetic Stripe reader. Then you are able to browse and book restaurants within that area. In the future you can also book activities in the area with your card, but I haven't gotten to that point yet (wasn't my plan to do so, but something I want to do in the future).

I also used sockets so we can use a single server with multiple Arduinos, and this reduces the access on the API and therefore we need to scrape less things, and the data is also stored centrally.

For the implementation I used NodeJS as that works amazingly with Arduino, and goes well with any input/output intensive system like this one. Database was MongoDB, and for the web I used AngularJS to serve the content to the user. In the future I could simply just use something like an LCD to display a website, but I don't think it's a pressing issue for me now.

The overall process is:

Arduino RFID/Button Press -> Streamed to the Website in some action -> Queries cache or API -> response given or performed

#final

Final Project: Overview

explain whether your project ended up as you initially envisioned

if not, how did the idea or implementation change?

Changed in a very dramatic way. Initially I thought I was solving a problem that I thought many people had, but it turned out that only I wanted to see that particular thing exist. I started off trying to create something for the food industry to make it easy to people to order and book seats easily, but it didn't seem like it was a problem for everyone else.

Then I pivoted to try and create some kind of agent where people can easily get food delivered to their houses even simpler to the ones that already exist. The ideas I came up with would just complicate what has already existed online today (Seamless, Grubhub, etc).

Then I moved towards create something that I have implemented now. My final idea was to try and get data from food websites to show people things as they approach different districts. They can easily see places they can eat, and other things within that particular district. The tool would give you food locations based on where you are now and give you the ability to book your food right away. Moreover, you can also use your metro card to identify yourself on the system (as most New York people have these).

what challenges did you face?

I faced a lot of technical challenges. There are a shortage of food APIs available for individuals to be able to use to gain this information. I had to manually go and scrape some of these websites to get information that should be provided easily.

Another challenge was the UI I had to create. I had to make it simple enough for individuals to be able to use on the street, and simple enough that anyone could use it quickly. I also couldn't use RFID's this time to identify people, but had to think outside the box to come up with a solution where everyone could use it - Tourist or non-tourist. The idea was to use their metro cards instead because most people have those.

how did you overcome them… or not?

I couldn't overcome the non-tourist solution yet. I haven't been able to secure a metro card reader (Magnetic). But that's going to be the next plan for me.

The API challenge was really hard for me. I built a tool that used PhantomJS to scrape seamless, and a bunch of other websites. It took the data from those websites, and put it in a API for people to use.

any lessons learned?

Designing a UI is really hard. Especially if you expect people to try it out in busy streets, and give them good results. Also to take into account Edge Cases that occur with people is a little hard. For example, if a person doesn't have a mobile phone - the app wouldn't work well because there is no way for a person to validate them selves (They can just browse, but not reserve).

There is also no keyboard because on the street this becomes a little hard for people to do. But I had to get around it using some methods. Touch would have been a good option, but I wouldn't be able to work with this in prototype phases (Time spent was mostly on software, and couldn't figure it out well enough by that time)

any last thoughts about this project?

Amazing learning experience. Used a NodeJS, PhantomJS technology stack, and got great results. Also used Socket.io so we can use multiple Arduinos to communicate and just use one server. Wasn't able to get a hold of a magnetic card reader (For the metro cards), but got the RFID working from the last project.

any last thoughts about physical computing?

Very interesting class. Got me thinking about a lot of product analysis, and idea generation. Also got me thinking about the cross of hardware and software, and allowed me to explore my passion for them further. Got a lot of knowledge out of the Labs, and the coursework.

I have also attached my powerpoint: http://www.scribd.com/doc/223334391/District

I have also attached my code: https://github.com/AbhiAgarwal/District

I have not attached the code for the scraper yet. I will make it available soon!

#final

Final project: Pivot

I made a pivot in my project today. I realized some loop holes after doing some usability testing with my friends and a few people in New York. I realized that this product would be a little silly in some instances, and really useful in others.

My new idea is to create a device that would be cost-free and could be deployed around New York in places that are touristy or get busy. It would treat as an interactive device that would help you figure out the particular area and give you recommendations. You could also use your past details to figure out your way, and find places that are good to eat and then drink after. The system would somewhat be merged onto the system I had previously where I had an RFID card.

I will be utilizing the details from my previous system but also adding things, and making it not for check-in systems this time.

#final

Execution and implementation (2 weeks)

I have built the client-side functionality of this. The user is now able to order/look through the food, and also look through a series of restaurants that are close by as well. Say if you're at a restaurant, but you're feeling like having something else then this would work as well.

The functionality I have yet to build is the arduino side. I haven't really understood the design fully for this yet or how I will lay out the buttons for the user or how the user will fully interact with the system. This is the last step for me to complete this project, and I will spend a while thinking about this before implementing it.

#final

Ideation and planning (1 week)

explain your idea in one paragraph

I want to create a simple way for individuals to get take out food when they are passing restaurants. It's usually very complicated if you're at the restaurants and you want to order take out - you've to wait for the staff to give you the menu, and sit and wait. This will allow you to use an interface of menus outside the restaurant or in a particular place and let you order food.

Also to add in this can make it easier to reserve or book restaurants or other things as well. A nice simple interface outside of a restaurant.

Suppose you're passing by a restaurant and you want to take a look at their menu and order the things right away. Then it would be awesome to just click a button, and it remembers your information and you load it up. I think this would be pretty great, and would give restaurants a good opportunity.

who is the target audience?

Explorers or people who usually walk around New York. When they see something they like they can order it and select where to order it at - either their homes or get it right there.

Restaurants who don't want to add features of take out but pickup, and if you're there already then you can pick it up on the time it tells you.

does your project solve an existing need?

I believe it's something I would enjoy having. Something that people can put outside their restaurants and people can take a peek and order food easily. Or reserve some seats - you sometimes have to call in, but if you're passing by why not have a nice interface.

does your project enhance an existing interaction or human behavior?

It decreases the amount of time you've to wait to order or get your food ready. You can simply focus on getting take out food or on the menu rather than having an individual wait for you to order or have bad customer service.

what parts do you need, how much do they cost, and when are they due to arrive?

They have arrived - the parts for the Arduino are a few nice interface buttons you can click to move around and order, and to navigate.

what, if any, hardware, mechanical devices, and/or movements will your project involve?

Navigation buttons, and a few instructions on how people are able to use this piece of technology.

what, if any, Arduino, Processing, or other software applications or libraries will your project use?

AngularJS, NodeJS, Arduino, and library to get the menu of a restaurant quickly

who are your team members

Just me!

what are their backgrounds?

Abhi - Computer Science background, and programming.

what are their majors?

Abhi - Gallatin - concentrating in "Engineering Creation"

what are they good at?

Abhi - Programming, and formulation of project

what is the division of labor?

I will personally be doing all the work!

#final

For this example I monitored the word "Justin Bieber" being mentioned on Twitter. For this I used twitter4j the Twitter Library for Java, and I used this particular tutorial to learn it. I used the same code I did for the Line Graph assignment, but this time I dynamically updated the count for each second, and put it into an array, which would reset every 1000s. So we could get a 1000 results before it would time out, and ran that in a huge while loop so every time the array reached 1000 it would just start back at 0.

twitter4j was extremely easy to plugin and use in my processing code, and the code was short as well. Then I was able to use the Twitter API docs to pull in some requests for that keyword and pulled the count each time. So every-time a new request came through I would increment the count for that particular second by one, and then just ran the analysis that occurred and outputted the graph as shown in the video above.

The general process was as shown below, and I had global array for count and would just increment the count depending on the result.

Twitter myTwitter;

void setup() {

myTwitter = new Twitter("..", "..");

try { Query query = new Query("Justin Bieber"); QueryResult result = myTwitter.search(query); } catch (TwitterException te) { println("Couldn't connect: " + te); }; };

#lab8

For my line graph I got data from the Bureau of Labor Statistics of the U.S. I compared the Labor productivity (output per hour) via the Percent change from previous quarter at annual rate for Non-farm Businesses, and Businesses. The line going linearly downwards is for Non-farm businesses, as a oppose to the other one which is for Businesses. It's interesting because the x-axis the the year, and y-axis is the change in productivity in the major sectors. My data was generally just the values below, and they are per year and not per quarter.

Non-Farm Businesses

Farm Businesses

#lab8

Visualizing a single series of data

I did mine on the Tuition History at New York University. My source was CollegeCalc.org, which is a calculator for tuition and fees for individuals in the U.S.

My dataset is the following:

$38765,$37372,$35290,$33420,$31690, $30094,$28496,$26646,$25380,$24336, $23456,$22586,$21730,$20756,$19748, $18739,$17640,$16650,$15620,$14502, $13336,$12250,$11130,$8714

Where it's going from 2010 to 1987. We can clearly see that there has been a positive rise. I manipulated some of the code I found on open-processing, and used it to create this particular bar graph.

#lab8

Control the speed of a toy motor from Processing

I'll upload the video for this particular project in a little while. My video isn't correctly uploading right now (Will upload when on NYU's network rather than home).

The code for this particular project is below. The code for this particular one is very similar to the code for the first section of this lab.

Arduino (Different to Lab 1, just using motorPin rather than Servos)

const int motorPin = 5; void setup(){ Serial.begin(9600); }

void loop(){ if(Serial.available() >0) { int message = Serial.read(); int motorVal= map(message, 0, 255, 0, 180); analogWrite(motorPin, motorVal); } }

Processing (The same code as Lab 1)

import processing.serial.*; Serial arduinoPort;

void setup(){ size(320, 240); background(0, 255, 255); println(Serial.list()); String portName = Serial.list()[1]; arduinoPort = new Serial(this, portName, 9600); }

void draw(){ int servoVal = (int)255*mouseX/width; arduinoPort.write(servoVal); println(servoVal); background(255, 255, 0); fill(255, 0, 0); ellipse(mouseX, mouseY, 50, 50); }

#lab7

Use a potentiometer to control the speed of a toy DC motor

I'll upload the video for this particular project in a little while. My video isn't correctly uploading right now.

The code for this particular project is below:

Arduino

const int motorPin = 5; const int potPin = A0;

void setup(){ pinMode (motorPin, OUTPUT); Serial.begin(9600); }

void loop(){ int potVal = analogRead(potPin); analogWrite(motorPin, (potVal * 0.20)); }

#lab7

Controlling a Servo Motor from a Processing Program

My apologies for the vibration - I realized after I switched over to the next project.

#lab7

Making Stuff - NYC ACRE

• Who owns/runs the space? - The NYC ACRE is run by a group of individuals, advisors, and investors. It's primarily based off of individuals who are interested in solving problems in the cleantech industry. It's the flagship cleantech incubator in New York City, and works with New York University, and the local government within New York City to fund and help incubate companies in cleantech. - The leadership is basically Micah Kotch - the director of Innovation & Entrepreneurship and Emily Wheeler - the deputy director. They are a couple of the biggest innovators in the sustainability sector, and big players in the New York startup scene. It's alumni currently consist of Ecological, Sollega, and current class includes Radiator Labs, and Voltaiq, which are two trending startups at this moment.

• Where is it located? - They are located in Brooklyn, specifically: 15 MetroTech Center, 19th floor, Brooklyn, NY 11201 They are walking distance away from the Polytechnic campus, and some of their class does include startups from NYU-Poly.

• What services do they provide? - They are an incubator, which means: they fund, advise, and give startups a place to grow and develop their cleantech applications. They also provide hardware/software assistance, and teams which are able to help them build out their infrastructure, and help them minimize their hardware costs. They enjoy helping students with their projects, but most of their resources are geared towards helping people who work out of their office.

• What maker tools do they offer? - They offer space, advisement, and resources that startups would need to create physical hardware for prototyping.

• how much does it cost to use their services/tools? - It's free of charge to apply to their program, but there are strings attached once your project/startup is registered. You can also use their office space, but that has to be applied for. Some are free, such as advising, meeting advisors & investors, and meeting some of the startups there.

• Are there discounts for NYU students? - It's a incubator - but they partly work with New York University, and NYU-Poly to work with some startups going through the NYU-Innovation incubator. The benefit of this is that some of their advisers and leadership team has overseen some of the most successful cleantech startups, and they know how to develop both software/hardware applications and how to market them. - There is a great push for them to promote their work to NYU students, and bring more NYU students into their field - so they give some preference to NYU-based startups.

• Photos/videos of the space and main tools Attached the photos!

• Photos/video of the owners/managers Couldn't take videos - there were people working when I went so I wasn't able to take videos. I am taking another visit at some point on a Saturday so I will definitely upload them then.

• Give a detailed overview of the steps involved in using one of their core tools. For example, if the tool is a 3d printer, describe the steps a fellow student would need to go through in order to have their idea converted to a 3d printed prototype at this location. - Advising: Email them and schedule time to meet an individual. - Working there: Apply to one of their startups, and contact them for an interview/ meet their workers. Or apply to the ACRE itself. - Prototyping: Go to one of their amazing workshops and learn about what they do, and the different hardware tools they have. To get to use their machines you've to put forth your idea, and some initial drawings and machines you would need. If you need to use the 3d printer you can fill out a form, and they would need to approve your request. Most of the time you wouldn't be able to because you're not apart of their incubator, but they make some exceptions for great projects, and great students. -- NYU also has a good relationship with them, which really helps.

• Give at least 3 ideas for how this location could be useful to a student project. - Advising: If a student is interested in learning and development a few projects based on cleantech they are able to find individuals who have just taken the step to registering their company and getting their first funding so they're a good resource. - Research: There are also individuals who are doing some incredible research so students are able to help them in their more theory based research and help them develop technologies. This is useful as it helps students help or lead research that they wouldn't normally have the opportunity to. - Initial Product Development: If individuals have ideas/projects that they want to prototype then they can come here. They have some state of the art equipment for hardware, and some great in-house libraries for software. So they're a great place to come and start your project with their help.

#makingstuff #writeup #lab6

Midterm Project Presentations

Introduce members of the team

• Name Abhi Agarwal • Major and any special interests Gallatin: "Engineering Creation", mixture of Computer Science (Artificial Intelligence), Mathematics, Engineering (Robotics) • What background in electronics or programming you each have Programming: Started programming pre-high school, and electronics: none, but I did 4 years of Physics through high-school. I've used many programmer languages in the past. • What were the roles in your team? I was the only one in the team, and I did all of the electronics & programming. • Did you divide labor and work mostly independently, or mostly as a collaborative? Alone in the team! • Was there an appointed or a de facto team “project manager”? Alone in the team!

Project Summary • Title Tap-and-checkin • Short pitch Usually it takes you a while to pull out your phone go on the specific app and checkin to that place, but we've come up with a solution where you use either your NFC equip Android phone or a RFID-chip to checkin to that place! Each location will have the device and a certain location ID and when you tap your card it will automatically check you in and save the hassle! • Does this fill a need? I believe so - because everything is being automated, and people are frequently checking in to let their friends know where they are or collect points - it could be useful! For example, in hotels people can tap their cards and let either the managers or their friends know that they are here. It's already being done by Foursquare, Yelp, Facebook, etc. and it simplifies the need.

It also can be used to checkin to multiple social networks. If people want to checkin to Facebook, Foursquare, Groupon, Yelp, and many other tools then you can use this instead as well! • Does this mediate or add to an existing behavior? Adds to existing behavior. We're just targeting people who are already using this kind of features. Also major markets are moving into data analysis, and knowing when your customer comes is the perfect way. • Is this a project you intended to be more than just a class project? Yes - I was thinking about this for a long time. I've been wanting to do this as I think people get frustrated of checking in because they don't want to waste their times taking out their phones and tapping them - it almost becomes like a hassle if you travel frequently so why not just tap!

The target audience • Who are you making this for? Anyone who already uses checkin apps such as Foursquare, Facebook, Yelp, Groupon, etc. • What tendencies and/or limitations does this audience have that you took into consideration when designing your project? The limitations are that these people have to have these existing services when trying to sign up. There is a lot of overhead required, and registration beforehand. Not everyone uses Facebook to checkin, but many people use Foursquare, and so targeting different users is difficult. • Did you do any user observation on your target audience when designing this project? Yes! I tried to make it as simple to combine your RFID number with your Facebook/Foursquare/etc, and that was really hard of a problem beforehand. • Did you test your project with your intended audience? Yes - I tried to use some people who had iPhones to use the tags instead and it worked pretty well! The Android people had some bugs though! • If yes, what did you learn? iPhone users had no problem with the tags, but Android users did. The Android users had some issues where the NFC ID changed very frequently and each tap led to a different hexadecimal number - so I worked around it and tried to build an app for a temporary solution and that fixed it. But the app isn't a solution as most people don't want to download another app in conjunction to their location apps - so I'll have to work with some Android Guidelines and try to parse the NFC content rather than the number itself. However, as it could make their lives easier some users wouldn't mind downloading the app.

The technical system • What hardware did you use? Arduino, RFID/NFC Scanner from Adafruit. • What software did you use? Arduino IDE to get and parse the RFID/NFC number, and sent that data to Node.JS (SerialPort in Node.JS) where I used OAuth (Node-Foursquare library) to checkin them to their service, and MongoDB (Mongoose in Node.JS) to store their OAuth key and RFID/NFC number.

Then I used websockets to control & show the login/registration/output on the web page to user. I used the library socket.io in Node.JS to implement this. The webserver underlying in Node.JS is implemented using Express. • Did you use any Arduino or Processing libraries? Yes - I used the RFID/NFC library from Adafruit for their particular software. • Any new types of circuits or electronic components you had to learn to use? I learnt how to quickly parse and use the RFID/NFC scanner, and how to parse through the different inputs it gives. • What aspect of developing the project took the most time? Parsing from the RFID/NFC scanner took a while, and then because the IDs were changing for the Android system it took me a while to write the app, which keeps it consistent. The MongoDB part also took some time as my logic wasn't quite right, but then I got that done as well!

Problems • What difficulties did you expect to have when you first started, and did they occur? The integration between Node.JS and Arduino was really simple, and took only a few minutes to setup correctly - so that was really positive. • What did you struggle the most in this project? The Arduino code for the RFID/NFC was a little intense - the library was really useful but I wanted to try get some of the data from the NFC out, and I still haven't been able to accomplish that. MongoDB parsing became quite a pain - the RFID numbers aren't always consistent and sometimes miss out on a few values so that become a little annoying. • How did you overcome these problems? I haven't yet! • Or why didn’t you, if you didn’t? I don't think this current reader gets the underlying data, but just the number Lessons learned • Was this project more or less work than you imagined at the beginning? It was about the same amount of work as I realized - I had to wait quite a few days for the RFID scanner to arrive so I started a little later than usual. • Did you have to learn any new software or hardware techniques to make this? I had to properly learn the RFID/NFC libraries, and also had to learn how to solder well. I also tried to use the WiFi shield so my device can be a stand alone! • Did you learn anything interesting from working on this project? Yes! I learnt a lot of good integration techniques between Arduino & Node.JS, which made it extremely simple to do things in the future.

The future • Do you think this project has a use outside of class? I think so! I would definitely use it in the future myself for my apartment and develop it further. It's quite simple and could be integrated into many things. Such as using the keycard to turn on your lights/house/door, and whatever else. • Do you intend to continue working on this? Yes! • Why, or why not? I would love to add some authentication techniques to other social networks, and let other users try it out in the future!

I have also linked my code on Github.

#midterm

Create an Etch-a-Sketch-type application using two variable resistors that can be used to draw on the screen.

Also an image:

I worked with Emily Dubovoy on this question.

#lab5

Oscillating Circle with Processing

What language is Processing based on?

Processing is built upon Java, but has simplified the language down an incredible amount - in terms of the syntax and the graphical visualizing tools. Rect and similar functions don't exist in pure Java.

In your own words, what is the difference between a global and a local variable?

A globally defined variable is accessible to an individual in every scope that is linked to that particular file. It can be accessed by functions/objects outside of the file, and also everywhere within the file itself.

A locally defined variable is accessible only within the function it is defined in, and within the scope of that particular function. It can't be used outside of the function itself, or depending on how large the scope is. Sometimes there can be private locally defined variables, which can only be accessed within that particular class definition or a function.

What data type does the draw() function return when it is called?

It returns a void datatype, which basically means it doesn't return anything back to the function that calls it.

In your own words, explain in detail what the code below the “//keep the circle in bounds” comment does

It stops the red ball from falling off the page. It has tight bounds so when it reaches those particular bounds it reduces the balls speed and it's position to go back in the negative direction. So when it falls to the bottom of the page it will go back up instead of just falling into an x,y position that is not visible, and does the same on the top half.

I worked with Emily Dubovoy on this question.

#lab5

Sending Serial Data from an Arduino Circuit to a Processing Application

No questions for this section

I worked with Emily Dubovoy on this question.

#lab5

My apologies for no videos this week - my internet was being crazily slow, and I have no idea why but it would take literally 3 hours to upload each video

Children's Toy - Fly it high

My toy was basically who can make a full circle the earliest with the time clock on the right. The puzzle is that you have to get 60 right buzzes for you to be able to move the clock the whole way around.

To get one buzz you have to press the Force Sensitive Resistor within 1 second of the Red Light Bulb going off. So the game is that the person who gets the most answers right within an x amount of time will get the points. To make this game even harder every time you don't press the Force-Sensitive Resistor and you miss a bulb within the second, you get timed out for 4 extra seconds.

Process:

Red Light Shines -> Hit Force Sensitive Resistor within One Second -> Clock/Timer on the bottom right moves 1 extra point

OR Red Light Shines -> Hit Force Sensitive Resistor after One second -> Clock/Timer does not move -> 4 second timeout

OR Red Light Shines -> Don't hit it within 15 seconds -> Timeout from the game and reset to 0 points

The point of the game would be to play with another individual and the person who's clock/timer on the bottom hand reaches the same position first - he would win!

#lab4 #toy

Trending Blogs

Recently Viewed Blogs

Physical Computing