SAW CONNECTED RADIO USB, NODE LIBARY SEND AND GET MESSAGES
NOOB WITH HACKRF INSTALLED
- My Name is Ali Payne
Things I have built
- MVP IOT system for fortune 500 Boating corporation to monitor and control sensors attached to boat with in collaboration with Technicity LLC based in Columbus Oh
- Mobile payment processor certified by Verifone that can process payment from phone through POS cash register from pump
- Launch Self-serve piloting campaign for Kroger to automate data science for best coupons to best customers
Why I am here
- Co founder and CTO of company by the name Adyptation
- Dayton, Oh
- helps people with chronic illness quickly identify health management strategies. Adyptation's software quantifies treatment effectiveness by capturing an individual’s biological rhythms during everyday activities in a non-obtrusive way. By providing an objective measure of the individual’s response, we facilitate the partnership between patients and providers in determining the best course of action.
Team Membership and Execution - Diversity of Team Members
How diverse are the team members’ experience levels, skills, and backgrounds?
CEO - Ryan Jankord
PHD in Biomedical Sciences
Professor @ Wright State University
5 years was Air Force Research Laboratory (AFRL) Director of Neurobiology of Cognitive Performance
COO - David Malseed
15 proven years of successful entrepreneurship
why are we here
- “We are funded by NSIN to autonomously monitor biomedical in austere environments and want to learn about military needs and technical requirements and know more about how we can add value to the military mission." -Ryan Jankord
- company recently completed SBIR Phase I Final Report for Proposal Number F183-004-1411
- Edge Processing: Computer processing currently relies on bulky and heavy hardware and power sources. How can we improve the ability to perform complex, analytical tasks (e.g., computer vision, natural language processing), in the field, without access to cloud computing?
- A real world example of edge processing would be the following
- I spoke with Alex Price and Benjamin Malone they represented 10th SFG and SOCOM
- They explained to me a situation in which they send out one individual to a remote area where they have no connection to internet they need to collect the biometric data on several individuals to build data profiles which can be uploaded. Right now just for taking iris scans and finger print they carry 20 pounds of bulky equipment
- I propose the solution of creating a uniform of raspberry pi all raspberry pi being connected in a mesh network communicating on the same radio frequency. Because all computers can communicate by messaging each other via radio wave lengths we can in a sense create a cluster of computers which we can use to process information or save information in a distributed manor
- First we need to verify we can send messages between two raspberry pis send data over fm radio using raspberry pi
- How much Data can be send via radio wavelengths
- QAM256 you can easily fit 50 Megabits per second into a 10mhz channel width
- Can we build cluster computer with rasberri pis?
- The power of the eight server CPUs (32 cores) will allow you to execute computations from the client CPU much faster than the client could perform them on its own
- Example used wifi router but it can be done
- Can we build a program that can take data convert it to binary and then send the data in packets to nodes in with cluster
- Small program in Python to go from Video to binary
'''python started = 0 hash_val = 0
with open("video", "rb") as file: byte = file.read(1) # read a byte (a single character in text) byte_val = ord(byte) # convert the string character into a number
if started == 0: hash_val = byte_val started = 1 hash_val = (hash_val << 5) - hash_val + byte_val # this is a basic hash
- Used python because it’s really pseudo code but can be implemented in multitude of different languages
- Send data to various sensors
Just need to take packet of data transferred to binary and create unique id and store along with the unique id for node that data was sent to
If we can glue these 5 possible steps together I know that we can solve this issue
- Data Transmission: Operators often have requirements to transmit large amounts of data to command elements when there is little or no communications infrastructure available. For example, they may be required to stream high definition, full-motion, real-time video or store up to 72 hours’ worth of data for later transmission. What innovative capabilities can be developed to collect, manage, and transmit data in austere environments?
A real world example of data Transmission would be the following
- I spoke with Thomas Laning who represented 5th SFG
- He explained to me main issue is that we have a bunch of people on ground need to be able to send live data 5, 10, or K miles away. Right now we are carrying a lot of batteries, large satellites, routers, and large computers
- Same as previously stated solution for first problem. Basically you first have to store the 72 hours of data that’s what problem one tackles now to send the data
- How. Much data is in for example 72 hour video clip
- format = MPEG-2-3?Mbps fixed rate
- resolution = 720 * 486
- frame rate = 23.98
- video length = 72 hours
- Total Space 117.07 GB
How much data can we send at a time
- Earlier we discovered QAM256 you can easily fit 50 Megabits per second into a 10mhz channel width
- So how many megabits are in 117.07 GB
- 936,560 megabits
- Divide by 50 because so that’s how many messages we have to send
- If we send that many messages on one radio signal that would be
- 18,731.2 seconds == 312 mins == 5.2 hours
- If we send that message with lets 500 radios
- 118,731 / 500 = 38 mins == half an hour
- Doesn’t take into account putting video back in order
- Every sensor would be responsible to sending individual packet of data
- 500 means 100 computers on 5 people or 5 computers on 100 people with
How much data can we send really and how far?
• Black hole of information would love to talk to subject matter expert
- Frequency hopping – Montor Bill Cohen
- Long story short we can speed thins process up
If system was implemented Other benefits
- Let's say we have solider with 5 mini computers in uniform mini computers can also intake sensory information like temp, motion detector, etc. Know that we have a way to store a lot of data we can collect it and all the data science people have a large data set to run whatever analytics where not then possible
- This would be ideal for triage medical they can receive data on soldiers
- If you have network of computers if you all take image you could create 3d replication of current situation
- Distributed systems are hard to spy they would need to know interworking’s of communication routing
What I learned
- What is a wireless mesh network?
What is fog computing(fog networking, fogging)
when to use fogging or edge processing
What is distributed network
What is Disputed computing
Full duplex vs half duplex sending radio signals
- In a distributed system if I separate data and then one of my nodes that has data is lets day destroyed I have no way of getting that information back so I cant guarantee you will also have all data
- To change this one would need set up lest say a pub sub network where one radio frequency can send a message to multiple minicomputer and we can duplicate to guarantee probability of not losing any packets of data.
- JAMMERS COULD SHUT DOWN SYSTEM
Impact - Problem Statement
- based upon the given situations I believe the real problem is what is the best way to have and IOT network without using the internet
Impact - Problem Alignment
- Fog IOT network
Impact - Need, DoD
- How significant is the DoD need for the proposed solution?
- SEEMS LIKE IT SOLVES THE ISSUES PRESENTED IN THE HACKATON SO IF THIS HACKATON IS BASED ON REAL ISSUES IN THE FIELD IT SEEMS AS THIS WOULD BE OF GREAT SIGNIFICANCE TO THE DOD
- What benefit does it provide over the status quo?
- The current solution is heavier
- The current solution has processing limit
- It’s the future
- Impact - Need, Private Sector
How significant is the private sector need for the proposed solution?
- What benefit does it provide over the status quo?
- In the private sector if you can create 3d landscapes of areas you can sell as a service
- You have a system that can collect data form sensors, which can be sold in the private sector to insurance companies, hospitals, schools, etc
- Having sensors and a way to collect data from them allows one to build a large data set that can be used for Health care proposes
Team Membership and Execution - Hack Journey
- How did you use this information to iterate to improved approaches? How well did you incorporate new information learned at the hackathon into your solution? How did you use this information to iterate to improved approaches?
- Radio vs satellite
- Antenna length send long distance message
- Distributed system mean more areas for failure you need to look out for
- Battery consumption
- Product/Service – Applicability # How well does the proposed solution work in anticipated operating conditions?
- Not really sure how a uniform of computers would hold up in the dessert but I think you could wrap minis with cooling gel and in winter weather it maybe a warm uniform
Product/Service - Practicality for Stakeholders
- How practical is the solution for all stakeholders (not just end users)?
- It’s a solution to a problem there are some pros andcons
- Its buildable
- Its more modular and customizable
- Product/Service - Resources required at deployment
How reasonable are the resources required to deploy the proposed solution?
• Product/Service - Technical Feasibility
How credible is the scientific basis for the proposed solution?
- We are just gluing together things that can already be done independently
Product/Service - Innovative approach
- Is your approach based on out-of-date technologies or ground-breaking ideas?
- Mixing a little old school with a little new school
Go-Forward Plan - Award potential
- With. $30k we can build an MVP product within 3 months
Go-Forward Plan - Post-award investment potential o See SBIR Phase I Final Report for Proposal Number F183-004-1411
Go-Forward Plan - Startup
- Is there opportunity for startup company formation for an individual team's solution or expansion of an existing startup team's solution market to include the Department of Defense?
- Adyptation would love to do more work with the Department of Defense
Go-Forward Plan - Future technical risk narrative