ChetanPatil Testing

Author: admin

  • PCBWay – Smart PCB Prototyping And Manufacturing

    PCBWay – Smart PCB Prototyping And Manufacturing

    Photo by PCBWay

    Developing an electronic product requires a lot of iteration. A hardware electronic device has to go through a detailed design process which is followed by numerous prototyping and evaluation. It is often a time-consuming and cost-intensive process.

    Not all the development processes happen in the house. The majority of the companies do in house designing and then rely on trusted partners to manufacture prototypes and later on the scale based on the outcome. It is critical to understand how one can design and where one can manufacture.

    MAJOR STEPS IN HARDWARE DEVELOPMENT

    There are two major steps before the hardware product hits the market: Designing and Prototyping

    Designing a hardware product for consumers requires a dedicated team that can provide proof of concept using different tools. After the idea is validated using circuit simulation tools, then the team will generate a detailed Bill of Materials (BOM) which will have all the details regarding components required to assemble and manufacture a hardware prototype in form of Printed Circuit Board (PCB). In most cases, the designing team will also provide Gerber files that will have the details of how many layers the PCB will have, which components are placed at what locations, orientation of the components, and many other assembly related details. All these technical details and information is then sent to the team which is capable of producing PCB for rapid prototyping.

    Prototyping is a very complex process and requires that the BOM being assembled adheres to industry standards. Majority of the prototyping happens using PCB which is outsourced. To assemble PCB, the service provider should have dedicated tools and teams that can take Gerber files, validate that it is correctly designed, and then process it through the manufacturing flow to provide faster PCB prototyping.

    WHERE TO MANUFACTURE

    There is no denying that China is the leader in PCB assembly with over 50% of the global market share. Shenzhen, which is often termed as The Silicon Valley of Hardware, is the go-to place when it comes to prototyping and large scale manufacturing of hardware products.

    Numerous companies provide elegant solutions in China and PCBWay is one of the leading manufacturers of PCB designing, prototype, fabrication, and assembly. It is one-stop solution for all things hardware manufacturing.

    PCBWay has opted a different approach that combines the power of software and hardware to provides manufacturing without compromising quality while ensuring timely delivery and cost effectiveness

    It is becoming a powerhouse of smart PCB prototyping and manufacturing.

    THE PROCESS

    PCBWay makes sit very easy for any customer to get started with the service. Traditional there are two types of design and development companies:

    • One which prefer to only outsource PCB and then do assembly either in a house or at other vendors
    • Second which wants everything to do done the same manufacturer: PCB, assembly, testing and scaling
    Photo By PCBWay

    For both such types of companies, PCBWay has solutions. Using PCBWay’s online quotation system designing company can opt for a quick quote. Apart from standard single-layer PCB, the company offers advanced multi-layer PCB designing and prototyping. Flexible electronics is one of the upcoming markets and PCBWay already has technology that can easily provide rapid prototyping and assembly of flexible PCB.

    All this is very handy for customers who would like to just get a quick prototype of PCB built based on the design provided, and later on, prefer to assemble somewhere else.

    If the OEM wants everything from PCB designing, manufacturing, and assembly to be done at a single location, then PCBWay already has all such resources. They can handle sample prototyping to turnkey production.

    Photo By PCBWay

    QUALITY

    It is very critical to ensure that any hardware being assembled goes through all the required quality checks. PCBWay assembly and manufacturing adhere to all such requirement that includes design rule check, automated optical inspection, electronic and probe test, automated X-Ray inspection, impedance control, RoHs lead-free, UL certification and different manufacturing tolerance.

    They have also partnered with leading design houses which ensures all these quality requirements are met from day one.

    COMMUNITY

    Apart from catering to the industry, one of the major steps PCBWay has taken is to differentiate itself from other manufacturers in China by engaging with electronic enthusiasts. They often run an online competition and PCBWay Community is one of the fastest-growing assembly communities.

    Anyone with or without knowledge about PCB manufacturing and assembly can sign up and learn rapid prototyping.

    PSA

    If you watch videos from Scotty Allen then you will definitely like below one by him on PCBWay.

  • Remote Monitoring And Management

    Remote Monitoring And Management

    Photo by Chris Liverani on Unsplash

    Monitoring of activities is not a new concept. Pre-computer era too human activities used to get monitored. With advent of high speed connectivity, monitoring has taken remote approach.

    Remote Patient Monitoring System alone is expected to be worth $1.8 Billion by 2026. Considering all possible areas which can make use of remote monitoring and management, the market size is expected to grow 10 to 100 folds. There are many areas which require remote monitoring like utilities, logistics, energy sector, weather research and many more.

    In order to understand and develop a robust Remote Monitoring And Management (RMM) solution. It is critical to capture different functions that needs to work together in order to make it more viable.

    Let us consider a practical health care solution where a patient has been released from hospital after recovering from COVID-19. With the help of RMM, a solution can be developed which ensures that if the symptoms of COVID-19 reappear on patients, then corrective actions are initiated.

    In order to build such RMM following key functions needs to to work in harmony:

    • Smart Hardware —- Data Generation And Acquisition
    • Smart Connectivity —- Data Communication And Monitoring
    • Smart Storage —- Data Storage And Access
    • Smart Software —- Data Analysis And Actions
    Photo By Chetan Arvind Patil

    Smart Hardware For Data Generation And Acquisition

    Hardware for patients RMM will require many components and the most critical are sensors. Sensors play critical role by generating and acquiring data.

    Sensors are capable of capturing activities like change in temperature, movement, heart rate, sweating, and many others. If at source the data is corrupted then any decision taken at later stage will not be accurate. Thus, it is vital to ensure that the accuracy of the sensors is within the acceptable range. Off course, there are other units like DSP, SoC, and WiFi for connectivity, which are also equally important and these come by default in any smart hardware.

    When data is captured, the next stage is to transfer it securely.

    Smart Connectivity For Data Communication And Monitoring

    As soon as the data is generated (with OR without COVID-19 symptoms) at source using smart hardware, then it needs to be transferred for long term storage and analysis. For application under study here, Bluetooth is the most practical solution due to low impact on energy consumption.

    This requires that the smart hardware with sensors is connected and is setup with mobile or desktop application. Sensor will periodically transmit the data using wireless connectivity. However, since connectivity comes with some downtime, the smart hardware will also require enough memory to store data at source. Then as soon as the connectivity is established, data can be transferred. After acknowledgement the memory from smart hardware can be cleared for future data.

    Other smart connectivity solution like WiFi, 3G, 4G and in future 5G can also find use in RMM only if the monitoring smart hardware has direct source of supply. For applications like patient monitoring, will require efficient battery management due to small form factor and thus Bluetooth is preferred compared to other.

    As a next block after the data has been transferred, it becomes important to focus on secure storage.

    Smart Storage For Data Storage And Access

    As soon as the data is available and is stored at the next immediate location (in this case a smartphone), it needs to be transmitted to a server. Once, the data reaches server then physicians can analyze and take actions.

    The critical piece here (Patient related RMM) is that when one applies technology to patients, then HIPPA comes into action. Which basically means that the data related to patients should be kept private, secure and much not be hackable.

    In order to achieve data security, as a first step end-to-end encryption is required by default. It is equally important to make the storage location secure and another layer of monitoring to capture intrusion should be implemented.

    As a final step, analysis of the data is done.

    Smart Software For Data Analysis And Actions

    It is important to understand that physicians are not data scientists. For a data scientist it is easy to play with raw data and take actions. When it comes to commercial application like the case study of patient RMM, physicians are looking for key factors related to specific disease to take corrective actions. Hence, it is important to present results such that any physician can understand it clearly.

    In most of the cases when patient RMM is developed then it is done in collaboration with physicians to ensure that the data presented follows the known medical terminologies.

    Only few of the RMM solutions will make use of existing data analytics tool in the market, but majority of the others will develop there own tools as that allows Over The Top (OTT) features that may allow other possible revenue streams.

    Take Away

    COVID-19 Patient RMM was an example on how RMM needs to be built using four key functions. Same functions can be applied to:

    • Manufacturing to track operations
    • Logistics to capture where about of goods
    • Utilities to understand consumption every second
    • Motor Vehicles to capture health of the vehicle
    • and many others.

    Many of the RMM solutions can be developed without the Smart Hardware, but only when the data to be monitored doesn’t need to be generated and is readily available.

    It will be interesting to see how RMM thrives along with Remote Working.

    Photo By Chetan Arvind Patil

    PSA

    Shirley Ryan AbilityLab in Chicago has been working closely with patients to develop many RMM. It has many useful resources for those interested in learning about RMM at the intersection of academia and industry.

  • Re-Imagining Algorithms For Outbreak Risk Management

    Re-Imagining Algorithms For Outbreak Risk Management

    Photo by Patrick Assalé on Unsplash

    In 2011, March Andreessen rightly pitched Why Software Is Eating The World. In 2018, I wrote a small note on Hardware Is Running The World. In 2020, algorithms play a crucial role in understanding and thereby catering to the needs of the world for many decades to come.

    These smart systems and context-aware algorithms are capable of capturing market trends, which has been made possible due to the amount of data that is generated every day. All these smart systems are also capable of making decisions for others and are often termed as Artificially Intelligent.

    During a COVID-19 like pandemic, AI systems could have made decisions in favor of the consumers, enterprise, and businesses. Smart systems catering to manufacturing, logistics, transport, eCommerce, and last-mile delivery would know very well that there is going to be a global demand for essential products based on the data world is sharing (as privacy is dead). That would mean more informed decisions were possible and could have driven the production and delivery of essential goods well in advance.

    Example: During COVID-19, a smart algorithm powering manufacturing industry could have projected future outcomes and provided insights to companies to ramp up the production of essential products. Which in turn could have helped adapt logistics and supply chain industry as per the needs of the cities and countries around the globe. This would have also lead to eCommerce to have enough essential products irrespective of the demand, and would have also enabled warehouses and stores to limit how many essential products a consumer is allowed to purchase well before COVID-19 hit the curve. Similar data-driven outbreak predictions can be applied to any industry, not just those producing essential products. 

    However, instead, it seems there was no projection of crises to come, and thus leading to shortage of essential products. 

    It can be questioned that outbreak based algorithms are already in place but are accessible only when one pays for it. However, given how every technological discussion is incomplete today without talking about data, isn’t it expected to have such a solution embedded in the data tools for any industry as default feature? Isn’t prediction the key for making smarter decisions?

    Focusing on the pandemic, there have been numerous attempts to make data available that could provide insight into coming epidemics:

    However, it seems with the COVID-19 crisis such data driven outbreak risk prediction solutions either failed OR were not utilized to full potential.

    Post-COVID-19, the data tool war is only going to get more intensive with the major focus on Data-Enabled Outbreak Risk Management. The important question will be whether these AI-Enabled algorithms are capable of making smart human-oriented decisions during crises.

    PSA:

    Kira Radinsky’s work on data based prediction is really interesting for anyone looking to read more technical details on prediction algorithms. She apparently wrote about importance of algorithms to predict the next outbreak in 2014.

  • Work From Anywhere

    Work From Anywhere

    Photo by Avi Richards on Unsplash

    There is no denying that COVID-19 is testing businesses based on how prepared they were to enable employees to Work From Anywhere. One of the key factors that drives Work From Anywhere culture is the resources an employee gets that allows him/her to be as productive (in many cases more productive than working from the office space) and connected as one would be when working from an office.

    Sure, there are companies whose business type doesn’t allow Work From Anywhere. For example warehouse management, restaurants, grocery stores and many more. But the major test for these types of businesses going forward is to adopt a business model that is more geared towards remote management that minimizes human interaction (without eliminating human jobs) and without sacrificing profitability.

    Businesses that can allow employees to Work From Anywhere also need to provide the right tools. Those who work at the software and hardware domain already have access to many tools and must have also observed that 90-95% of the tools they use can also be accessed from anywhere.

    The post-COVID-19 world is going to be slightly different. It might just give free hand to employees as to whether:

    • They want to work a day shift
    • They want to work a night shift
    • They want to work any shift from anywhere

    If the answer is any shift from anywhere, then one also needs the right tools all the time. These tools are a combination of software, hardware, and connectivity. Software and hardware are provided to employees by default in by any company that encourages Work From Anywhere. Also, if one looks at the way of working in software and hardware domains then most of the tools are at the remote locations that can be easily accessed with the click of a button.

    Below is the list of a few things that I believe every (not just high-tech companies that can afford)
    business needs to provide their employees with:

    Hardware

    This is 101 if companies want their employees to Work From Anywhere. Without the right hardware with enough computing resources, it will be meaningless to expect employees to be productive. Given how affordable many of the smart devices are, this should not be a big issue for many companies. For many whose, companies by default provide laptops to work with might take it for granted by considering that same is applicable for other companies too. But given that providing hard assets to 1000+ employees also require a dedicated IT team to manage, it makes it more difficult for many companies to let everyone have such take home assets. This raises the need for affordable asset management. It is not that all employees should get laptops. If the work responsibility is different, then companies can move from laptops to pads to smartphones. It all boils down to what hardware resources are right for the job the employee is expected to do.

    Software

    Right hardware also needs the right software. If you are in hardware designing, you need the right software that is mainly living on the remote server. If you are in software development, you need the right IDE to work that and it may or may not need remote execution. If you are working in customer support, you need telecommunication tools. For doctors, they need patient management and diagnosis tools. For the restaurant business, it can be efficient delivery of goods.

    Connectivity

    Whether the resources the employee is accessing requires VPN or not, by default VPN should be a must and that also means need for robust internet connection. This enables secure connection and minimizes the risk of hacking. As of today, it seems many employees are paying out of the pocket of connecting the tools to start Work From Anywhere, but it seems like paid healthcare future employees may also demand Paid Connectivity as a requirement in their offers.

    For those who work in the tech industry all this seems like “we are already used to this”, but one thing many people don’t understand is when millions of people lose their job due to the pandemic, it also raises the question about how many of these millions of people can work remotely and still be an asset to the company.

    In the end it is all about right people working anywhere with the right tools.

    Anecdote:

    A few days ago, I had to get my account details with the internet provider updated. To do so, I called the customer support and a lady answered my call. 5 minutes into the call while she was in the middle of taking care of my request, I heard the background voice of a baby crying. The lady promptly apologized for it and I was quick to tell her that I understand.

    This made me think strongly: Has it always been that customer support employees were Working From Anywhere OR it is just COVID-19 that has made more and more people work remotely OR is it just that this particular employee signed up with the internet provider for remote assistance service due to lack of resources during COVID-19? Something to wonder about.

    It all sounds like a cliché, but this is how it’s going to be in post-COVID-19 world: Work From Anywhere.

  • The Last Enemy – Total Information Awareness

    The Last Enemy – Total Information Awareness

    Photo by ev on Unsplash

    There is no denying that all online activities are being tracked. Either by the governments, or by companies. Data is the future oil, and the more companies have it the better their chances of survival.

    There are several factors as to why it is so easy for companies to get the data and make money out of it. Below are the reasons that I think contributes to easy leakage of data.

    People Do Not Care About Privacy:

    • This is true, people use technology for convenience and 99% of the products which provide solutions that customers (We The People) desperately needs, then they are willing to give up their privacy.
    • On top, if it’s free and provides with solutions that everyone needs then privacy word doesn’t matter.
    • There are tons of examples of where this scenario is applicable.

    Products Can Not Be Trusted:

    • All the software out in the market: You can’t trust them when it comes to privacy and tracking
    • Real Life Scenario:
      • On my smartphone, location is disabled
      • Specific apps don’t have permission to use location when not being used
      • Many times after I park my car and walk out, I get a notification “Why worry about parking hassle when you can take Lyft”
      • This just shows how I simply cannot trust both the hardware (smartphone) and software (apps)
      • Also, there is no way to solve this other that using a brick phone

    Data Logs:

    • Delete words doesn’t exists and doesn’t opposite of what it should in software data world
    • Google Maps, without an argument the most used location app and by default it knows where you have been, how long you have been, what route you took where you parked, whose home you went to etc.
    • Google does provide options to disable logging of such information.
    • However, can we really trust such features? Does delete really means deleted because it’s not visible to us?
    • What’s if after deletion of data, it’s still retained forever? Facebook does.

    Privacy Tools Do Not Work:

    • One of the most preached tool to protect privacy and fight tracking (after encryption) is VPN.
    • Problem with VPN is that it’s like giving your data after paying for a service.
    • Yes, arguments can be made that VPNs don’t log your data because they say so.
    • However, this is simply not true as VPNs may save you from hackers trying to steal data but VPNs provides in no way can gurantee they don’t use your data to make money out of it.
    • This basically means (at least to me) that VPNs are surveillance tools too, just that it’s one that people pay for.

    All the above helps create Total Information Awareness systems that companies and governments can build easily without much effort. It will be fair to say that any product you use, doesn’t matter whether it is hardware or software, is in itself a Total Information Awareness tool.

    If you are more curious about privacy and mass surveillance, then do watch The Last Enemy which showed how mass surveillance works in modern era well before Edward Snowden went public.

  • Moral Machines

    Moral Machines

    Photo by Denys Nevozhai on Unsplash

    In 2018, Uber self-driving car under test in Tempe, Arizona was involved in a crash which unfortunately leads to the killing of a pedestrian. Last week, National Transport Safety Board concluded that it was Uber’s self-driving software’s fault (apart from various non-technical valid issues), as the autonomous software was not programmed to react to pedestrians crossing the street outside of designated crosswalks. This flaw (which Uber seems to have fixed now) raises a question about situations in which software, when not programmed correctly, can lead to more severe crashes.

    This reminded me of Moral Machine, a project at Massachusetts Institute of Technology, that creates extreme scenarios (similar to trolley problem) to understand human perception. The data collected points to the fact that every individual has a different perspective to the same extreme situations.

    This is interesting, as self-driving cars are designed and programmed to do what humans have been doing for over a century: driving the car. If people have a different perspective about a hypothetical crash situation, then how will an autonomous car react to such situations? How does software considers this?

    For sure, the programmers writing code for autonomous cars are smart enough to take all this into account, but with Uber’s technical flaw it surely means that moral machine concepts cannot be overlooked. There will be scenarios when the software will follow specific rules, and that may be safe.

    Moral machines concept is something to think about, as the industry is still far away from providing technological solutions that will make self-driving cars hardware and software do what human brains can.

  • Crowd Sourced Private Mass Surveillance

    Crowd Sourced Private Mass Surveillance

    Photo by Bernard Hermant on Unsplash

    Crowd sourcing is not a new concept and the term was coined around 2005. In nutshell, it allows individuals to participate and complete tasks that are part of a bigger project. Contributing crowd may do it voluntarily or get paid for it. The concept of crowd sourcing is great, as long as it’s used for good cause. For example:

    However, if crowd sourcing leads to surveillance that too a private one then one should start questioning whether the intentions are good or bad. I came across one such example after landing on Cyberwire podcast and it opened many other details that were new to me.

    The podcast talks about how companies like Digital Recognition Network (DRN) (and many others) are using crowd sourcing to capture every license plate out on the roads. DRN has created a surveillance database of 9 billion license plate scans. More than the population of the world. This all is possible due to ubiquitous cameras, participation due to crowd sourcing and Automatic License Plate Recognition (LPR) libraries like OpenALPR, an automatic number-plate recognition library.

    Technological advancement is good as long as they are used for good intentions like: Finding stolen cars, car recalls, locating cars for finance recovery, amber alerts, unregistered cars on road etc.

    However, things start to go out of hand when same database starts acting like a surveillance tool and that too for private usage. Yes, cars are everywhere out in open, but that doesn’t mean anyone (except authorized authorities like Police etc. with a cause) should be able to simply enter license plate number to find where all this particular car has been during its life. The worse that can happen is such data base is exposed or hacked and then linked to vehicle registration databases, eventually creating an Open Graph about all car owners.

    Another example of crowd sourcing private surveillance is digital doorbells, where all the owners intentionally or unintentionally are contributing to the data base that allows private policing.

    The major reason to write about this was to look at the other side of it and raise concern about Crowed Sourced Private Mass Surveillance.

    Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety.

    Benjamin Franklin

  • Power, Performance, And Energy Management of Heterogeneous Architectures

    Many core modern multiprocessor systems-on-chip offers tremendous power and performance optimization opportunities by tuning thousands of potential voltage, frequency and core configurations. Applications running on these architectures are becoming increasingly complex. As the basic building blocks, which make up the application, change during runtime, different configurations may become optimal with respect to power, performance or other metrics. Identifying the optimal configuration at runtime is a daunting task due to a large number of workloads and configurations. Therefore, there is a strong need to evaluate the metrics of interest as a function of the supported configurations. This thesis focuses on two different types of modern multiprocessor systems-on-chip (SoC): Mobile heterogeneous systems and tile based Intel Xeon Phi architecture. For mobile heterogeneous systems, this thesis presents a novel methodology that can accurately instrument different types of applications with specific performance monitoring calls. These calls provide a rich set of performance statistics at a basic block level while the application runs on the target platform. The target architecture used for this work (Odroid XU3) is capable of running at 4940 different frequency and core combinations. With the help of instrumented application vast amount of characterization data is collected that provides details about performance, power and CPU state at every instrumented basic block across 19 different types of applications. The vast amount of data collected has enabled two runtime schemes. The first work provides a methodology to find optimal configurations in heterogeneous architecture using classifiers and demonstrates an average increase of 93%, 81% and 6% in performance per watt compared to the interactive, ondemand and powersave governors, respectively. The second work using same data shows a novel imitation learning framework for dynamically controlling the type, number, and the frequencies of active cores to achieve an average of 109% PPW improvement compared to the default governors. This work also presents how to accurately profile tile based Intel Xeon Phi architecture while training different types of neural networks using open image dataset on deep learning framework. The data collected allows deep exploratory analysis. It also showcases how different hardware parameters affect performance of Xeon Phi.

  • Dynamic Resource Management of Heterogeneous Mobile Platforms via Imitation Learning.

    The complexity of heterogeneous mobile platforms is growing at a rate faster than our ability to manage them optimally at runtime. For example, state-of-the-art systems-on-chip (SoCs) enable controlling the type (Big/Little), number, and frequency of active cores. Managing these platforms becomes challenging with the increase in the type, number, and supported frequency levels of the cores. However, existing solutions used in mobile platforms still rely on simple heuristics based on the utilization of cores. This paper presents a novel and practical imitation learning (IL) framework for dynamically controlling the type (Big/Little), number, and the frequencies of active cores in heterogeneous mobile processors. We present efficient approaches for constructing an Oracle policy to optimize different objective functions, such as energy and performance per Watt (PPW). The Oracle policies enable us to design low-overhead power management policies that achieve near-optimal performance matching the Oracle. Experiments on a commercial platform with 19 benchmarks show on an average 101% PPW improvement compared to the default interactive governor.

  • Semiconductor Fabrication

    Semiconductor Fabrication

    Photo by Mathew Schwartz on Unsplash.

    Samsung’s technical blog has series of articles on semiconductor fabrication. It covers majors steps from tape out to packaging. All nine parts of the series are not linked together, so I thought of creating a list that may help those interested in learning about semiconductor manufacturing.

    All the images below are from respective part of the series linked in the title.

    Part 1: Creating the Wafer

    Picture By Samsung

    Part 2: The Oxidation Process

    Picture By Samsung

    Part 3: The Integrated Circuit

    Picture By Samsung

    Part 4: Drawing Structures in Nano-Scale

    Picture By Samsung

    Part 5: Etching A Circuit Pattern

    Picture By Samsung

    Part 6: The Addition of Electrical Properties

    Picture By Samsung

    Part 7: The Metal Interconnect

    Picture By Samsung

    Part 8: Electrical Die Sorting (EDS)

    Picture By Samsung

    Part 9: Packaging and Package Testing

    Picture By Samsung