When we commit to developing products from the ground up...... We get our team to work in order to get a prototype off the napkin and in the customer's hands — how do we do that? Well, We’ve chosen to invest in rapid prototyping because it allows us to create tangible products from our computer-aided (CAD) files, as opposed to just 2D drawings, with a relatively quick turnaround time. The major stages we cover include:
- The product requirements document (PRD)
- The prototype components and the process for creating each
- Planning the project budget and timeline accordingly
These are based on many years of installation, repairs, creating custom-made fittings, including prototypes, and a few years of working with designs. Most customers always want to know the “ballpark figure” or the cost of their project after briefly telling us about their idea. The reality is that product development work involves all sorts of design — simple and complex. Because no two projects are alike, therefore a detailed product requirements document (PRD) is required to get the process started and, of course, to even get a quote.The PRD sets the tone for how the product will be designed and manufactured. We like to think of it as a list of the non-negotiables of the product. It’s in the PRD that we include the following:
- Your main product application — where it’ll be used, how and by whom.
- The minimum or maximum size the customer would like the product to be.
- How long the battery should last (think of the use cases for your product)
- Whether or not it should be waterproof, weather resistant, fireproof, etc.
- Accuracy of any feature the customer may have (the more accurate they want something to perform, the more expensive it gets).
- Any integrations — how it should work with other systems.
Let’s say a customer has commissioned a custom-build standalone or remote smartlock system. A lot of time will be spent picking parts with a lower power consumption and optimizing it for a longer battery life. For a device like an indoor digital display, on the other hand, a customer wouldn’t really be concerned with battery life since it will have to be plug into the wall socket. These are the types of items that are well documented in the PRD. Our design team then gathers the information in the PRD and start designing the product around these constraints — along with what will be realistic based on a customer's budget and available technologies. We take the information in the PRD and make into a statement of work (SOW), which will outline project-specific activities, timelines, deliverables and other payment-related information. The last important thing in the PRD is that it’ll cover each of the components of a prototype.
Most hardware products we design usually consist of four different parts:
1. An enclosure made of plastic, metal or another material
2. A printed circuit board or other electronic components
3. Firmware, or the code that runs on the electronic device
4. Software, or the code that runs on the computer or phone to interact with the newly developed hardware
However, some prototypes don’t always have all of these parts. For instance, a remote control case would only have a plastic component and none of the others would matter. Or maybe a customer wants to launch an educational development kit (such as the Arduino), so there may be no really need for a case or software. In any case, The hardware enclosure design is usually a two-step process.
The first step is to have an industrial designer sketch out several different concepts of what the product could look like based on what it’ll be used for. The sketch is often done by hand but our industrial designers use software. Several sketches may be required in order to ensure that the hardware enclosure not only meets a customer's vision but is feasible for manufacturing. We usually do four or five sketches to gather feedback before settling on something the customer really wants. If a customer is still in the initial steps of product development and only wants a minimum viable product (MVP) or a rough prototype to analyze its use cases, we may skip refining the drawing and just work from an initial sketch. At this point, functionality is more important than aesthetics and adding industrial design to the mix would just be more costly and lengthy for someone who just wants it for testing purposes.
After the sketch is finalized, a CAD designer will create a model for the prototype using SolidWorks, AutoCAD Inventor, Pro Engineer or Catia. It’s in this step that the engineer will specify tolerances, fittings, assemblies, DFM (design for manufacturability) features, etc. The 3D drawings usually take a lot of time to model. It can take three to four weeks or several months depending on the complexity of the product.
The end result will be a set of different types of files usually composed of 3D CAD files, 2D drawings for manufacturing, and any bill of materials (BOM) for off-the-shelf parts that go into the product. The actual 3D printing or CNC of the prototype, based on these files, will happen after the design phase is complete and then the next phase begins which is the design of the circuit board, or the brains of the product, also often follows a two-step process.
The first step is the research and development of the product. Some of the products we work on are very innovative and have never been done before. That means it’s hard for us to estimate how long it would take to prove out the concept before trying to design it. The second step is a proof of concept (POC). Working in uncharted territory usually results in a proof of concept that will test your product’s function and technology — but it’ll look nothing like the final product as we’ll most likely use breadboards and off-the-shelf electronic parts. The POC's only purpose is to make sure that the product idea is doable using what is technologically feasible at the present moment. We’ll use breadboards, microcontrollers, sensors, jumper wires and other electronic components in order to get a POC.
Yet, a lot of our products skip the POC if the product involves technology we know will work. If that’s the case, we’ll usually jump straight to the circuit board design. We always do the circuit board design in tandem with the enclosure design, unless the product does not have an enclosure. Unlike the enclosure design process, however, the circuit board design normally takes longer to make once the first design is ready. This is because the circuit board prototyping process is very slow. Not only do we have make a POC, but once it’s done we have to design the PCB on the computer, make the bare circuit board, create a stencil for the board and order all the components before we actually assemble the final pre-production prototype and test to ensure the design works.
It’s not like we can just go to a 3D printer and hit “print.” Every time we have an error we need to address, we have to make a new version, make a new bare board, populate it, program it, test it, rinse and repeat. This process can take between one and three months.
At the end of the circuit board design stage, we end up with a Gerber File that shows the schematics of the board and a Bill of Materials with the components that will be populated into the board. Most of the time we also deliver one finished, assembled prototype that the customer can now test in its environment before going into production or making changes.
Then the firmware which is software that gives the product life. To design it, we basically have to convert the product requirements into code. For example, if a customer want a blue light to come on when the device is connected, we need to program that feature. Turning on an LED is easy, but add 50 other things that need to happen in a multitude of scenarios, that is when we run into some pretty complex problems. And most programmers will know that these firmware design problems can take forever to figure out......Estimate several months if not a year to work out all the firmware issues. We usually program everything in C but have also written in C++, Python and many other languages.
The last part of many modern hardware products is the software, a program that allows the hardware to send and receive data over a connection while displaying it in a usable way. This is usually a program that runs on the computer — or on the web — or an application that runs on a phone. Fitbit, for example, uses a wristband with a microcontroller, accelerometer and battery to send the step count to the phone. The application on the phone converts this data into useful information. By the way, the application can be used to configure how often the band reports information to the phone to conserve battery life. But not all products require a software application to work — think of a Bluetooth speaker. If that’s the case, they only need firmware development. We would probably spend even more time on software development than we would on firmware development, mostly due to user interface (UI) and user experience (UX) design. Although they may bring hefty costs, investing in UI/UX will certainly help any application in being more effective, user-friendly and aesthetically pleasing — especially if the software is the part where the end-user will mostly be interacting with the product.
Now comes the most asked question: how much should the customer budget for the hardware prototype design? Well, Let’s say customer wanted a digital display and they could spend X, Y or Z amount for the full product design. All of these would get them a digital display prototype, but the difference between them would be the earth and moon apart. In the end, the more funding a customer has for rapid prototyping, the more engineering time we can spend on it and the better the product can get. With that said, we encourage customers to set their prototyping budget so that our design team can help optimize it to get the best results because we are all
about disruptive technologies.
Re-Purposing Products
The first step is the research and development of the product. Some of the products we work on are very innovative and have never been done before. That means it’s hard for us to estimate how long it would take to prove out the concept before trying to design it. The second step is a proof of concept (POC). Working in uncharted territory usually results in a proof of concept that will test your product’s function and technology — but it’ll look nothing like the final product as we’ll most likely use breadboards and off-the-shelf electronic parts. The POC's only purpose is to make sure that the product idea is doable using what is technologically feasible at the present moment. We’ll use breadboards, microcontrollers, sensors, jumper wires and other electronic components in order to get a POC.
Yet, a lot of our products skip the POC if the product involves technology we know will work. If that’s the case, we’ll usually jump straight to the circuit board design. We always do the circuit board design in tandem with the enclosure design, unless the product does not have an enclosure. Unlike the enclosure design process, however, the circuit board design normally takes longer to make once the first design is ready. This is because the circuit board prototyping process is very slow. Not only do we have make a POC, but once it’s done we have to design the PCB on the computer, make the bare circuit board, create a stencil for the board and order all the components before we actually assemble the final pre-production prototype and test to ensure the design works.
It’s not like we can just go to a 3D printer and hit “print.” Every time we have an error we need to address, we have to make a new version, make a new bare board, populate it, program it, test it, rinse and repeat. This process can take between one and three months.
At the end of the circuit board design stage, we end up with a Gerber File that shows the schematics of the board and a Bill of Materials with the components that will be populated into the board. Most of the time we also deliver one finished, assembled prototype that the customer can now test in its environment before going into production or making changes.
Then the firmware which is software that gives the product life. To design it, we basically have to convert the product requirements into code. For example, if a customer want a blue light to come on when the device is connected, we need to program that feature. Turning on an LED is easy, but add 50 other things that need to happen in a multitude of scenarios, that is when we run into some pretty complex problems. And most programmers will know that these firmware design problems can take forever to figure out......Estimate several months if not a year to work out all the firmware issues. We usually program everything in C but have also written in C++, Python and many other languages.
The last part of many modern hardware products is the software, a program that allows the hardware to send and receive data over a connection while displaying it in a usable way. This is usually a program that runs on the computer — or on the web — or an application that runs on a phone. Fitbit, for example, uses a wristband with a microcontroller, accelerometer and battery to send the step count to the phone. The application on the phone converts this data into useful information. By the way, the application can be used to configure how often the band reports information to the phone to conserve battery life. But not all products require a software application to work — think of a Bluetooth speaker. If that’s the case, they only need firmware development. We would probably spend even more time on software development than we would on firmware development, mostly due to user interface (UI) and user experience (UX) design. Although they may bring hefty costs, investing in UI/UX will certainly help any application in being more effective, user-friendly and aesthetically pleasing — especially if the software is the part where the end-user will mostly be interacting with the product.
Now comes the most asked question: how much should the customer budget for the hardware prototype design? Well, Let’s say customer wanted a digital display and they could spend X, Y or Z amount for the full product design. All of these would get them a digital display prototype, but the difference between them would be the earth and moon apart. In the end, the more funding a customer has for rapid prototyping, the more engineering time we can spend on it and the better the product can get. With that said, we encourage customers to set their prototyping budget so that our design team can help optimize it to get the best results because we are all
about disruptive technologies.
Re-Purposing Products
When clients upgrading their routers, we became creative...... What we do with these old routers? In the case of clients switching ISP, they’ll often be asked to return the older device. But when a spare router is kicking around the place, we have several ways to reuse/re-purpose these devices.
1. Wifi Repeater
When the Wi-Fi network doesn’t extend across the full range of facility..... Although might might opt for powerline Ethernet adapters, adding a second router into the mix becomes a good alternative. This means connecting the old router to the new wireless network, using the Wi-Fi signal. It can then share access to the Wi-Fi network, giving greater coverage. Although there may be some latency issues, overall this is a quick and easy way to extend your wireless network. It has various uses, from giving better Wi-Fi access to a remote part of the facility, to letting clients stream video/cctv footage,etc to tablet while they are within the facility.
2. Guest Wi-Fi Connection
When clients are having people regularly dropping in and they allow them to use their wireless network, why not give those people their own network? This is like the wireless repeater project, but with a twist. The router connects to your existing, password-protected network, but gives password-free access to new devices. This will use the guest network feature of your old router, which will by default prevent guests accessing other devices on your network. If this level of security isn’t enough, we then address the firewall settings on the main router to adjust.
3. Wifi Radio Streamer
When clients want to enjoy their own radio stations on their wireless network? Some routers can be configured to play internet radio, we install the OpenWrt or DD-WRT custom router firmware. Some other software is also required, and we’ll also need a USB soundcard to output audio. While this isn’t an easy build, and plenty of other internet radio options are available, this is still a great project. It gives insight into the power of custom firmware, as well as an appreciation of how music is streamed across the wireless network. However, we have build one without a fuss using our Raspberry Pi smart streaming speaker project which is a good option.
4. Network Switch
Most routers don’t have more than six Ethernet ports. With the increase in wireless technology, this figure might even be as low as four. But with a clear need for devices to be connected over Ethernet, we often run out of ports. For example, home appliance monitoring devices, TV decoders with smart TV functionality, games consoles, and more might have no wireless networking. They need a physical connection to a network, and that means Ethernet. We add network switches when we run out of Ethernet ports. This is basically the Ethernet version of a mains power bar, with the additional ports plugged into one port on the router. The old router typically has four or more ports, so connecting will instantly increase the number of ports available. We also power up the old router and also disable wireless networking on the old router, to avoid conflicts.
5. Wireless Bridge
When a new router is wireless only? Even when the ISP didn’t offer a router with Ethernet ports, or a client uses a cellphone to connect to the internet but requires an Ethernet port. Either way, when there is a need to connect Ethernet devices to the network, a wireless bridge is the answer. While inexpensive, an old router can be repurposed as a wireless bridge. This works a little like a wireless repeater, but rather than share the Wi-Fi connection, the wireless bridge offers Ethernet connectivity. The old router is then connected to an existing Wi-Fi network, and its Ethernet ports used to connect devices requiring Ethernet.
6. Smart Home/Office Hub
Some routers have some useful additional ports. In some cases, this might be a USB port, which makes flashing OpenWRT or DD-WRT router firmware easy. Other devices might come with a serial port, and these routers can be repurposed as a home automation server. Basically, the router runs a web server that is connected with an internet browser. This might be on a PC, or for convenience, through a smartphone. We use this with an Arduino (Microcontroller) hooked up to the router, and some RF-controlled power switches, to create a basic smart home/office setup. While easier options are available, we use this to get a better understanding of home automation in our workshop.
7. NAS Drive Router
When clients are looking for ways to store data on a single storage device and access it from anywhere...... We offer them Network Attached Storage (NAS), which is basically a hard disk drive that is attached to a network. While our NAS devices are affordable enough, with an old router hanging around, we continue to save money. Note that this is limited to routers that can run custom firmware (like DD-WRT) and a spare USB port, and routers that allows the browsing of the contents from any connected USB devices. Without USB port, there’s no way to attach the hard disk drive or USB flash storage. Once set up, our now custom-built NAS gives the client instant access to important data from anywhere using any device.....These are some of the great ways to repurpose old routers, and even if a router is really old and misses some key modern wireless networking features, we can still use it as a switch, or even a guest network.
(c) 2018, MEP Digital Systems (Pty) Ltd.
No comments:
Post a Comment