Author: Daniele Fusi

  • 7 things to consider when creating an IoT device on a small budget

    7 things to consider when creating an IoT device on a small budget

    TL;DR

    When creating an IoT device on a small budget, it is important to consider the following:

    1. Clearly define what you want to build and document it in a Product Requirements Document (PRD)
    2. Ensure the product solves an actual problem and conduct basic market research
    3. Stick to a Minimum Viable Product (MVP) and avoid feature creep
    4. Have a total budget in mind and plan for non-recurring engineering (NRE) costs
    5. Utilize off-the-shelf components and open source hardware

    What do you actually want to build?

    The question is deceptively simple. Many of us have a tendency to dive head-first into work and get blindsided by issues we didn’t even know existed.

    Engineers may lack awareness of business challenges, while business-oriented people may have a hard time understanding technical challenges.

    Creating the product’s most important document is paramount: the PRD (Product Requirement Document). This document defines what your product looks like and which features it has. This document must be easy to read and understand.

    As an example for a requirement written in a PRD:

    ”The device should run on battery for at least 7 days” is understandable by the whole team, while “The battery capacity should be at least 1,200mAh” makes the requirement hard to understand for non-engineers.

    Normally, the PRD will be a living document in the form of a spreadsheet.

    The simple one below is a starting point for designing a smartwatch:

    Product Requirement document example

    Similar to SMART goals, PRD requirements must be Specific, Measurable, Attainable, Realistic, and of course Time-based.

    Check out Mastering the Requirements Process: Getting Requirements Right, which tackles this topic in detail.

    If you work with other people, it’s essential that the latest PRD is readily available to everyone on the team. Nothing breaks trust like keeping secrets (yes, we’ve seen managers hiding the PRD from their own employees).

    Ideally, once this document is made, it is set in stone. In reality, this rarely happens.

    Changes in the PRD along the way result in unpredictable delays on the development process and increases in cost.

    Better to sit down, take some time, and define a robust, encompassing PRD. This document should go through few revisions rather than being edited every day.

    A strong PRD is your first tool in keeping the project on track. Time spent on it will pay you dividends.

    Look around – does this product already exist?

    Legal issues could be hiding in plain sight, if you don’t play your cards well. Make sure the product name can easily be trademarked, and nobody can accuse you of plagiarizing their existing device.

    If a similar product does exist, look for inspiration: what do people dig about it? What does it miss?

    An opportunity might be right around the corner.

    Does this product solve an actual problem?

    Much has been written about Silicon Valley’s talent in designing stuff nobody needs. Juicero and Washboard come to mind, and those are only a few examples.

    While one could argue that “People don’t know what they want until you show it to them,” we believe that line is best reserved for geniuses like Steve Jobs. For us mortals, instead, some basic market research is a must.

    Look around yourself for the pains people go through. Can you remove that pain from their life?

    Shipping something is better than nothing – Stick to the MVP

    An MVP (Minimum Viable Product) for a hardware device is a version of the product that has the minimum set of features necessary to satisfy early customers and provide feedback for future development.

    (from Scott M. Graffius, Agile Scrum: Your Quick Start Guide with Step-by-Step Instructions)

    The MVP is simply the list of requirements that are “must-haves” or highest priority in your PRD.

    An example of an MVP for a smart home device could be a device that can control the lighting in a room using a mobile app. The MVP would include:

    • A physical device (e.g. a small box) that can be installed in a room

    • The ability to connect to a home’s WiFi network

    • A mobile app that allows the user to turn the lights on and off and adjust the brightness

    • The ability to integrate with other smart home devices such as a voice assistant

    It’s important to note that MVPs are designed to be simple and minimal, so there may be additional features that are not included in the MVP but could be added later, in a new product version.

    When you’re new at product design, it’s easy to catch the feature creep fever: adding features your user doesn’t need.

    Instead, hold on to your horses and stick to the MVP you already identified. One of our partners recently made the choice of not having wireless connectivity in their first product. It felt bold to many on the team, but we knew it was the right choice. Adding connectivity could have meant not shipping at all (which is what usually happens). This way, instead, their users get the product they need, and in turn will provide feedback to inform the next product version.

    “Great products do less, but better.” – Fabricio Teixeira, Designer at Work & Co, Founder of UX Collective

    Total budget

    Even a simple IoT device can easily cost you $50,000 in NRE (non-recurring engineering) costs. Add to that the cost of the hardware itself, and you better plan for a six-figure budget.

    All of this budget pressure can be lightened by producing smaller batches at first and limiting features to the MVP.

    Working with small clients, we normally drive toward a first prototype that can be completed on a small budget. That proto proves that the initial idea was solid and helps raise more investment.

    Because we care about delivery as much as our clients do, we often push back when asked to add all the bells and whistles that will slow down progress.

    For example, Bluetooth connectivity will cost you ~ $8,000 in fees even if you sell one single device.

    Skills needed – what does it take to go ahead and build?

    If you got to this point, chances are you have some of the skills required to build and ship a device. Maybe you’re great at writing code, or you know how to analyze the business side of things.

    Either way, there’ll be areas where you need help. The following is a brief, incomplete Skills List

    • Hardware design and prototyping: Knowledge of microcontrollers, sensors, and communication protocols such as Wi-Fi or Bluetooth.

    • Software development: Proficiency in programming languages such as C, C++, Python, or JavaScript for writing firmware and software for the device and connecting it to the internet.

    • Cloud computing: Knowledge of cloud platforms such as AWS, Azure, or Google Cloud Platform for storing and analyzing data collected by the device.

    • Project management: Ability to manage the development process and coordinate the work of a team.

    • Business: Understanding of the target market and the value that the IoT product can provide to customers.

    Basic first proto

    To reconnect with the idea of MVP: your first proto will look messy and embarrassing. And that’s OK.

    Breadboard with diodes

    The goal here is to prove all the parts of your device can play well together, even if they’re not yet in a form that can be sold to a paying customer.

    Don’t waste your time polishing up the details – first, build a simple proto that works.

  • How to use Lithium batteries – and not blow them up

    How to use Lithium batteries – and not blow them up

    Written in collaboration with Bryan McLaughling.

    TL;DR

    • Lithium batteries are widely used in modern devices such as laptops, smartphones, and electric vehicles due to their high energy density, lightweight, and long cycle life.
    • It is important to charge lithium batteries properly to avoid safety risks and to maintain battery health.
    • When using a lithium battery in a device, consider the placement of the battery and ensure that there is enough space for expansion when fully charged.
    • The battery management system (BMS) takes care of the battery, including the charger, fuel gauge, and thermistor.
    • Lithium batteries should be treated with care to avoid issues and ensure proper function.

    What are they and why they’re so popular

    You probably have a number of lithium batteries lying around you right now. Most modern devices, such as laptops, tablets, smartphones, smartwatches, portable speakers, headsets, mice, keyboards, smoke detectors, and even your Tesla car contain lithium batteries.

    All of these products wouldn’t exist in their current form if it weren’t for lithium batteries.

    Lithium batteries are rechargeable batteries that use lithium ions as the main source of energy. They are lightweight, have a high energy density, and a long cycle life compared to other battery chemistries, making them popular in portable electronics, electric vehicles, and energy storage systems.

    High energy density is the key advantage here. You wouldn’t carry your smartphone around if it weighed 10lb (4.5kg) or was the size of your laptop. Lithium batteries effectively enable phones that can fit in your pocket and be carried anywhere you want. After all, lithium is the lightest of all metals!

    How to take care of them

    “With great power, comes great responsibility” – Attributed to a lot of people

    Only a few years ago, Samsung was forced to recall their Note 7 out of concern for exploding batteries. Consider that 2.5 million Note 7s were sold, and you get an idea of the financial impact on the company, plus the PR nightmare.

    Teslas and e-bikes have been catching fire too, raising questions about the control electronics that keep these batteries safe. Recently, a single e-bike battery started a fire that consumed a whole supermarket in New York City.

    How to charge them – safely

    Unlike old-school lead-based batteries, which can be found in ICE cars, lithium batteries need to be charged with care. The charging circuit will employ what’s called a “charge profile”, similar to the one below.

    Essentially, a dedicated charging circuit will first determine the battery status. In case the battery is deeply discharged, a proper pre-conditioning phase is a must to establish battery health. 

    Assuming the battery is healthy, charging can start at a high current level. This phase is called CC (constant current). Once the battery is ~60% charged, charging current will gradually decrease until the battery is deemed fully charged. This last phase is called CV (constant voltage), because the circuit will simply maintain the battery voltage constant while letting the current decrease exponentially.

    This is why it comparatively takes a long time to charge from 90 to 100%: when the battery is almost full, the charging current is very low. Charging from 0 to 10% happens much more quickly, because the current is much higher.

    To be fair, one could theoretically charge at a low, constant current for the whole charge cycle. However, that’d result in a much longer charge time. Most smartphone users would find that unacceptable.

    When it comes to picking a charger IC for your lithium battery, we recommend industry leaders like Texas Instruments, Microchip, NXP, Analog Devices or Maxim Electronics.

    Where to put them in your product

    Should you decide to power your device with a lithium battery, pay attention to a few key details.

    Where will the battery be placed? In most smartphones, for example, the battery is secured in the back, usually by means of strong double-side adhesive.

    Double-check that the battery has enough space to expand when fully charged. Yes, lithium batteries will grow bigger when at 100% capacity. You can expect them to swell up to 10% of their initial thickness.

    A mechanical engineer working with batteries must remember to include swell into their TA (tolerance analysis) and make sure they will have enough room around them. Otherwise, the device itself will deform when the battery swells, or worse yet the battery will suffer damage.

    On the electrical side, there’s a whole other slew of challenges to keep in mind. We’re looking at them in detail in the next chapter.

    Suffice to say that these batteries need to be treated with care, or else things will go awry.

    Common control electronics – BMS

    Open up any smartphone, and you’ll likely find components that take care of the battery itself. As a whole, that’s the BMS (Battery Management System) and it’s usually made of:

    • Charger. This is the main chip that takes care of charging the battery (following the proper charge profile we already discussed). It requires an external source of power, which often comes in the form of a USB cable from a wall wart. The charger will make sure that power is always available to the smartphone, whether the battery is charging or not.
    • Fuel gauge. Quite a fancy name, suggesting a combustion engine. Here the fuel is simply electric current. This chip specializes in measuring it to a high degree of accuracy. Often, the fuel gauge will “talk” with the charger as well, ensuring full monitoring of every charge and discharge cycle. Over time, the fuel gauge will get to know the battery and monitor key parameters such as SoH (State of Health) and SoC (State of Charge). As the battery degrades after years of use, the fuel gauge will take note and keep it going for as long as possible.
    • Thermistor. A simple component, placed right on the battery that tracks its temperature. Battery temperature increases during charge and discharge, which is a normal phenomenon. Ideally, temperature should stay within 60 – 95 F (15-35 C). However, this range can stretch 20 F on each side and you’re still in a safe zone. Anything outside of that, and it’s better not to charge the battery. Discharge at low temperatures can still happen, however it will drain the battery much faster than it normally would.

    Where do they come from?

    Lithium is available, though it isn’t exactly easy to extract. Concerns about mining and metal availability should play into your supply chain strategy, should you need lithium batteries in high volumes.

    Tesla went as far as considering mining its own lithium, given how much material they need to build their cars. We don’t expect Fusion Engineering Group going to that length in the near future.