Earlier this year I was asked by the team at ProdFest to come to the Product Festival in San Francisco and give a presentation as well as Q&A in regards to the importance of prototyping. The timing of the event was opportune as I had begun to sow the seeds of my transition from Breadware, the EE and FW engineering firm I started in 2015, to Ioterra (not yet branded at the time). As this was one of my first public speaking engagements, I was nervous... but also excited as I am truly passionate about the importance of prototyping!
The overall theme was that hardware development is a wholly different beast than software development and that prototyping is just as important (although executed differently) for both.
Although it would be great to approach hardware prototyping in a methodology similar to Eric Ries’s The Lean Startup, it is difficult to follow the same methodology with hardware simply due to the fact that you cannot “recompile” a physical object (the above GIF showcases elements of hardware development that cannot be compiled). Rapid prototyping for hardware products has come to the rescue of product makers in the recent past but it is still hard to iterate on a hardware product quickly.
It simply takes more time and money to build hardware prototypes. In fact, most companies only get one chance to go to production and ship an IoT product. This necessitates more planning for hardware prototyping, as it is more difficult to go through the “Build, Measure, Learn” loop.
In order to compile the best advice for the attendees of ProdFest SF, I approached my colleagues from the development industry and they were kind enough to share some of their hard-won knowledge and tips on prototyping with me. These were the main takeaways:
Since IoT ecosystem development involves hardware, software and firmware components, it is a complicated project. Therefore, rapid prototyping holds greater importance here. In my presentation, I shared the scope of contemporary techniques used for IoT prototyping.
3D printing for example is the most common type of rapid prototyping. Here, the development components are integrated and solidified through the control of 3D printing programs. The end output is a layer-by-layer assembling into a three-dimensional CAD object model.
It is then saved into an STL file format that further works as a source of reference for producing the actual output. Next, using a CNC Machining the computer monitored tools like the end mills or drills the solid block is modified into a custom pattern. This is helpful in producing multiple instances of the prototype. Moreover, multiple cuts of different angles can be made.
I personally recommend breaking down the prototyping at the following different levels:
Prototyping of IoT begins with mechanics. IoT development is 70% mechanical development (hardware + electrical) and the prototyping efforts largely depend upon the quality and the relativity of the product body. In IoT, mechanics is many things. So all embedded components in an IoT device are first proofed for relevance in prototyping. PCB (Printed Circuit Boards), sensors, LED indicators, wireless connectivity and all other components that are a part of the design are covered here. For mechanical prototyping, there are many ready-to-use launchpads. For more complex builds, I always recommend outsourcing to an IoT development partner.
UI- prototyping is performed by building a front system only to test the successful data population on screen and the user ease in reading it. Based on the feedback, the developers evaluate the ease of data access at the outermost layer of the product. UI prototyping involves a team of end users who test all the features of the product and judge it for its convenience. At the same time, it is a testimony of the technologies used here. In most scenarios, the latest technologies such as HTML versions, Vue, React etc. should be used. Moreover, a mix of HTML/CSS/JS is also a great pick.
Now, this is the most crucial part of IoT development. Backend development drives the business logic of the product landscape. Since there is consistent data flowing in & out of the system, big data management is highly important here. Therefore, the backend prototyping should focus more on the ability to capture, analyze and populate data in real-time and with no fail.
Some of the recommended programming languages here include JavaScript, Python, Ruby and PHP.
Ultimately, it is the seamless functionality check of the backend systems with UI. To ensure interoperability of the data flow from the backend to the UI, the end-to-end functionality check at the prototyping level is essential. While the UI-backend connectivity is performed with REST APIs, the device-backed connectivity can be done through IPV6 Meshing, Hub, Bluetooth, or Low Power Wide Area Networks.
In a world thriving with contactless communications, the demand for IoT development is rising every day. Consequently, the enterprises are facing production issues to meet the demands. Therefore, the accuracy of the prototype becomes even more imperative to ensure a qualitative product by the end of the phase.
The overall theme was that hardware development is a wholly different beast than software development and that prototyping is just as important (although executed differently) for both.
Although it would be great to approach hardware prototyping in a methodology similar to Eric Ries’s The Lean Startup, it is difficult to follow the same methodology with hardware simply due to the fact that you cannot “recompile” a physical object (the above GIF showcases elements of hardware development that cannot be compiled). Rapid prototyping for hardware products has come to the rescue of product makers in the recent past but it is still hard to iterate on a hardware product quickly.
It simply takes more time and money to build hardware prototypes. In fact, most companies only get one chance to go to production and ship an IoT product. This necessitates more planning for hardware prototyping, as it is more difficult to go through the “Build, Measure, Learn” loop.
In order to compile the best advice for the attendees of ProdFest SF, I approached my colleagues from the development industry and they were kind enough to share some of their hard-won knowledge and tips on prototyping with me. These were the main takeaways:
- “Build looks-like models to generate powerful feedback & excitement” - Ryan Olson, Stel Design worked with Inboard early on to build "Looks-like" models, allowing them to validate mechanics, construction, and function to generate anticipation and commitment among the community. Looks-like models are a great way to validate key aesthetic requirements in addition to generating market excitement.
- “Prototype using real data as often as possible” Johan Olsson, Topp Design , a Swedish design and innovation agency, was commissioned by Stockholm County Council to design their new public transport experience including ticket purchase, journey planning and traffic information. By creating prototypes that were connected to a real back end, the city could conduct real user research in the subway with the actual app users, very early on.
- “Merge the lines between engineering and manufacturing” Sam Miller, Ansync design often works on complicated electromechanical projects that require tight integration between design and engineering. They recommend maintaining an 80% IP portfolio (don’t recreate the wheel if you don’t have to), using next-day services such as Digikey and McMaster-Carr, and having your designers work in the same room as the manufacturer.
- “Speak the same language” James Lua, Priority Designs worked alongside Include Fitness to build a complex IoT digital health & wellness ecosystem that pairs HIPAA compliant cloud software, health informatics sensors, and inclusive equipment. Priority Designs’ team of researchers, UX/UI designers, mechanical engineers, software engineers, electrical engineers and prototype specialists had to collaborate and their success was greatly dependent on the communication between these groups. James recommends getting all the stakeholders involved from the beginning and dropping the jargon as the first step towards successful end-to-end IoT product development.
- “Test product positioning with the target market prior to launch” John Stump, GoChair John Stump worked with Rain Factory to implement a research and product positioning process to identify the target market and the ideal price point of the GoChair product prior to launch. Doing this allowed John to verify that the margins and cost structures were in line with what the market was expecting to maximize sales and ensure a profitable campaign.
Since IoT ecosystem development involves hardware, software and firmware components, it is a complicated project. Therefore, rapid prototyping holds greater importance here. In my presentation, I shared the scope of contemporary techniques used for IoT prototyping.
3D printing for example is the most common type of rapid prototyping. Here, the development components are integrated and solidified through the control of 3D printing programs. The end output is a layer-by-layer assembling into a three-dimensional CAD object model.
It is then saved into an STL file format that further works as a source of reference for producing the actual output. Next, using a CNC Machining the computer monitored tools like the end mills or drills the solid block is modified into a custom pattern. This is helpful in producing multiple instances of the prototype. Moreover, multiple cuts of different angles can be made.
I personally recommend breaking down the prototyping at the following different levels:
- Mechanical Prototyping
- User Interface (UI) Prototyping
- Backend system development
- Interoperability
Mechanical Prototyping
Prototyping of IoT begins with mechanics. IoT development is 70% mechanical development (hardware + electrical) and the prototyping efforts largely depend upon the quality and the relativity of the product body. In IoT, mechanics is many things. So all embedded components in an IoT device are first proofed for relevance in prototyping. PCB (Printed Circuit Boards), sensors, LED indicators, wireless connectivity and all other components that are a part of the design are covered here. For mechanical prototyping, there are many ready-to-use launchpads. For more complex builds, I always recommend outsourcing to an IoT development partner.
UI Prototyping
UI- prototyping is performed by building a front system only to test the successful data population on screen and the user ease in reading it. Based on the feedback, the developers evaluate the ease of data access at the outermost layer of the product. UI prototyping involves a team of end users who test all the features of the product and judge it for its convenience. At the same time, it is a testimony of the technologies used here. In most scenarios, the latest technologies such as HTML versions, Vue, React etc. should be used. Moreover, a mix of HTML/CSS/JS is also a great pick.
Backend Prototyping
Now, this is the most crucial part of IoT development. Backend development drives the business logic of the product landscape. Since there is consistent data flowing in & out of the system, big data management is highly important here. Therefore, the backend prototyping should focus more on the ability to capture, analyze and populate data in real-time and with no fail.
Some of the recommended programming languages here include JavaScript, Python, Ruby and PHP.
Connectivity to ensure interoperability
Ultimately, it is the seamless functionality check of the backend systems with UI. To ensure interoperability of the data flow from the backend to the UI, the end-to-end functionality check at the prototyping level is essential. While the UI-backend connectivity is performed with REST APIs, the device-backed connectivity can be done through IPV6 Meshing, Hub, Bluetooth, or Low Power Wide Area Networks.
Conclusion
In a world thriving with contactless communications, the demand for IoT development is rising every day. Consequently, the enterprises are facing production issues to meet the demands. Therefore, the accuracy of the prototype becomes even more imperative to ensure a qualitative product by the end of the phase.
