Blog

Invisible expenses – Hardware startup

We have seen a lot of people ripping of electronic items to see how much each component costs and then come up with a figure of the value of the item. We would like to weigh in a few expenses which we have identified and experienced apart from the cost of components, which is mostly ignored by many, so that it might help someone stepping into this field.

Assembling Cost: The PCB components are to be assembled. Some people need PCB Stencil, which costs a few thousands in rupees. There are also different finishes of PCB, all the way from least expensive leaded finish to expensive finishes such as ENIG. We mostly use ENIG in our boards for better finishes and quality. The assembling cost varies with quantity.

Transportation: All components, boards, boxes, cases, moulds, stencil, prototypes, cables, screws, batteries, papers, stickers and almost everything is transported and it costs a lot of expense. And an additional 18% GST (+1% cess if from states like Kerala) when billed from the vendor.

Moulds: These are the one time investments for making plastic cases of electronic items. These costs all the way from 50K for small moulds and upwards to tens of lakhs depending on number of pieces per mould to complexity, cooling, finishes etc.

Screws: Might not seem an expensive item, but is usually more expensive than the passive items used in electronics.

Seller platform fees: All online platforms charges fees, in various tabs, depending on the category, all the way up to mostly 20-50% including the shipping and handling charges (including GST). And this amount they collect is credited a few weeks to months later, depending on platforms. If you have your own site and set up a payment gateway, you might have to pay about 2-5% on transaction charges and 18% GST on it. Also be ready to shell out money on packing and printing labels and on shipping charges.

Printing charges: You end up printing a lot of sheets for shipments and invoices, not to mention other documents and accounting related papers.

Storage charges: If you make more, you need to have a bigger rented space to store more. This can’t just be any place as your product might wear out and be covered in dust if not stored properly.

Return charges: If you plan to get online to sell, be ready to accept returns. There are users who order, and return without even receiving the items, and others who order and return without reading the descriptions or looking at the images. Also, we observed that, only a few still contact the startup for support. Most rely on the easy return conditions of e-commerce platforms and do a hassle-free return than try to sort out the issues. In case of return, the seller (you) will be charged the shipping charges, and will also end up losing the packing, invoice and related expenses. The returned items will take and eternity to come back to you.

Items with issues: No machinery is 100% perfect. If you make a 1000 pieces, there will be items which are not fit for sale. This happens in all stages, from PCB manufacturing, assembly, components, final assembly etc. This is an overhead cost.

Salary & Rent: Now, this is mostly the avoided part. Everyone works for a salary and each stage requires a human input, which is to be paid for. Also, there is a rent to be paid for storage and other infrastructure.

Taxes: GST on all payments made. Additional cess for certain states.

Import duty: The items which you see on international websites can get up to twice as costly when it gets to your hand, including the GST, shipping charges, import duties, cess, surcharges and then handling charges, storage, import clearing charges, duty advancement charges and GST on top of these (except for items whose GST has to be paid at the time of import)

R&D: Development takes a lot of time, and lot of trials before it can hit the market or ramp up into production. There are small batches of components required and this is quite expensive (as much as 40x) depending on the quantity that you do. For instance, a prototype assembly for a board (30x50mm, with 25 components) would cost you, Rs. 500+GST for assembly (excluding the 2-4K you spend on stencils). The same production version might cost you <Rs.50 including the stencil charges and taxes.

Banking charges: Companies have current accounts, which doesn’t pay interests, and are charged per transaction.

Capital investments: There are equipment which you need to have to build the firm up, from soldering iron to oscilloscopes or even much expensive items, depending on need.

Electricity charges and related: On the commercial scale.

Government registrations, renewals, taxes and audits: There are charges for all these, and these are not limited to these mentioned tabs.

We will be adding valid points as we catch up with them, or if you can help us with something which had not been mentioned, kindly comment below.

RDOne: Prototype; RD03: beta launch!

Our first project which was on hold, the #RDOne is in the prototype stage again. A simpler low cost low power version of RDOne was already complete by March 2020.

After our launch of MicroUPS in June 2020, we had received feedback from customers and well-wishers. We also had our first experience on online platforms such as Amazon and Flipkart. From these valuable feedback and experiences, we have decided to move into a more premium segment. RDOne is thus updated to offer a more premium look and features. RDOne has also enhanced protections and almost everything better compared to current offerings in market.

#RD03 will be a modular version of the RDOne and will be intended at customers who want an economical and modular solution, with low financial investments. RD03 is currently in low volume production and will be available in market as a beta release for customers who would like to get hands on the product before the production line kicks in, hopefully by August first week. Beta version, will be available on our Web store initially, and depending on demand and request, possibly expanded to other online market places. Beta version will also be priced slightly less compared to the production model.

Uninterrupted Power Supply (UPS) for your Arduino / Embedded IoT project

Most of the developers or people who tinker with the Arduino boards, often find it difficult to find a proper power source. Most applications are either battery powered, or are always plugged in. These systems some times, suffer interruptions as their power source run out.

We had recently introduced our 5V UPS MicroUPS ( https://www.regaldreamtech.com/micro-ups/ ) for 5V devices. [The old version has been discontinued. A new version is available at https://hashwhy.com/5v-microups ] We had received a lot of inquires whether MicroUPS can be used for IoT projects on Arduino for power backup. The simple answer is, ‘Yes!’.

MicroUPS is a 5V UPS with MicroUSB B input and USB output port. Arduino UNO, for instance, requires 5V input. MicroUPS can be used in applications that require a continuous power back up. For example, if your Arduino project is a temperature sensor that always monitors the temperature or humidity using its sensors, the arduino might not work in the absence of main power. MicroUPS can ensure that the power supply remains in-tact when the main supply is absent, by switching to the battery power. Battery will recharge once the main supply is available again, and the Arduino will be powered from the mains when this happens.

A block diagram for the usage of MicroUPS for embedded DIY projects that need uninterrupted power supply

Not only Arduino, most embedded boards which needs up to 10W power on 5V supply voltage may be used such as ESP8266, Raspberry Pi boards ( < 2A versions ) and other embedded boards. MicroUPS comes with 5000 mAh BIS certified batteries which is mostly sufficient for common applications.

You can buy MicroUPS from Amazon.in or from our web store directly.

Webstore: https://hashwhy.com/5v-microups/
Amazon: https://www.amazon.in/dp/B09NC1PZXW

The first ever product launch – The journey

I would love to first share the journey, in brief, and then answer the most common queries that people have asked me, towards the end of this blog post.

To, start with, the company was registered officially on 9th November 2018, as a private limited company, in Kerala. I joined the company in Jan 2019, hoping that I would launch a few quick production run projects and pursue higher studies in a few months. As expected as it was, things didn’t go as planned. Our initial designs were not stable, and when they were, we found it difficult to source the components. Since the components were sourced in small quantities, preserving the cash flow, the price was quite high, with high lead times. We struggled to get hands on components which we needed for first few months, and the projects dragged on.

After a few months, our boards were good as we started assembling it ourselves, without relying on third party service providers, who were expensive, with massive shipment delays. There were standard items, such as the ICs and resistors which are manufactured for all, and are bought by us, and there are customized items as well. We procured such customized items from local manufacturers, since we wanted to save time and money in the first few runs. Now, these customized items were not up to our design expectations, or as promised by the vendor. This rendered our designs useless in extreme conditions, though they were working under normal conditions.

We learnt that, it was quite easy to customize our design to use the things that are already available in market, than to make customized packets as these are expensive. The vendors are usually reluctant to give out small quantities as well. We designed RDTwo, the MicroUPS, in June 2019. The designs were complete. We did a few test boards, and those were good. We wanted to be a bootstrap startup than absorbing the capital to expand. So, we planned for scaling from 25 pieces to 100 pieces and then to 1000 pieces.

Component costs were high, injection moulding way too expensive, and the assembly houses reluctant. After a few months of work, where we bought our own 3D printer and printed almost 100 pieces, and sanded down all the marks from pieces with hours of work, we knew this was not going to work out. The cost was way too less, the investment in terms of cash, even less compared to normal scaling, but the work that we had to do for this, was way too high. We also understood that, though we were able to manufacture about 100 pieces a week, this was way too low for production.

We moved to injection moulding and 1000 pieces production by December 2019. Things went as planned, until the Covid-19 hit. We had planned to launch the product by March end 2020. Our consignments were stuck in major cities as the country was locked down. Without much raw materials available, the work didn’t progress much. Even after the lock being relaxed, logistics didn’t resume though the government had permitted. We took consignments from major cities where the courier offices were filled with undelivered items for months.

And then, we slowly launched on 13th June, 2020, in for a level 2 of the startup.

Moving towards the most common questions that we are being asked.

Do we manufacture in India?
Yes. We import the main components from US/Singapore. Rest all works are within the country. Since we don’t have sufficient investment to get ourselves a PCB assembly unit, we give them to third party service providers in Kerala. We assemble the product ourselves, each part, including the stickers on the boxes.

Was this designed in-house?
Yes, all designs were done by us. Though the mould designs for our case was done by a third party vendor.

How much time have you invested?
A lot of time, energy, money and relations.

Which do you think is the most difficult part as of now?
Ramping into production for the first time, without proper experience.

Were you helped by people?
Yes, a lot! Friends, family and people whom I have never met or talked to before.

Is there a lot of pressure?
Yes.

More questions will be added as they are asked.

Thank you for reading!
~John

Baby Steps of Object Detection

Computer vision has been a hot topic in the past decade and this as well, with machines being able to identify objects as humans, with some errors as well. Since the pandemic has invaded the world, the world has chosen to stay still with minimal interactions to avoid the spread and infections. This has in fact affected us a lot, being a hardware startup. We had our product ready to be shipped when this hit, and with all revenues dried, combined with the lack of supply of components, we thought of stepping into the world of machine learning which required minimal hardware components.

Google Inc has come up with TensorFlow, a platform for neural network training. We, have initially used ImageAI on Python for our initial testing. We coded, and then found that the codes were running fine and the machine was learning after a few failed attempts. The most difficult part, we though, would be to find a proper hardware to test the models and to train them.

We opted to create a library of the things we wanted the model to identify and started with it. We tried training the model from scratch. But then, we later learnt that there are pre trained models which we could use and then train them for our data set. This turned out to be faster.

We came across Google CoLab while we had to run the models to train for hours. Google CoLab is a free tool which can run these training with specialized GPU and TPUs. We mounted our Google Drive for data set and to save the trained models.

This is our journey so far. Stay tuned, as we move ahead.

The tools that we have used are
https://imageai.readthedocs.io/
https://colab.research.google.com/

A snip from our training

Hardware startup ecosystem – The difficulties

One of the main drawbacks of smaller companies not easily getting into the product phase in India is the delay encountered and the lack of proper ecosystem !

From my personal experience, compared to a place like Shenzen (China) where there are readily available stocks of components, and facilities, India lacks a proper ecosystem. In a proper ecosystem, it takes few hours to a few days for iterations and designs, whereas, for us, it takes at least a week for the components to arrive, and then a few more days for one iteration of design !

Also to be shared, is the reluctance of the vendors to deal with small orders, and also the lack of proper expertise in their fields, though they have been working in the field for years, or at least claim to be so. When placing a small order, especially for the prototypes, most vendors either charge you very high price, or may even not reply to your request as their margins are much lower on your order.

On an average, the probability of getting a positive reply is 1/5, and that of getting the item you searched for at an affordable rate may be about 1/10. So, to the people who are tired of trying, you should be tying harder !

Project RDOne : Prototype level case & Brief overview of project

Most people, the well-wishers we have, have always asked about what we make, and how big it would be ! People also ask how much did it cost, how much will it cost and so on !

  • We don’t have much answers about the pricing as of now, we expect it to be placed at a maximum of Rs. 1000 / piece.
  • It is expected to fit into 40 mm x 40 mm x 30 mm cuboid without the cables, which will be permanently attached to the device.
  • The device would come only in black matte finish colour.
  • The complete case would be 3D printed for the initial lot or until the next version.
  • Initial 100 pieces of the production will be sold at a discounted rate, and upon the feedback received, updated and put into production at the normal pricing. The initial pieces is expected come with extended replacement warranty compared to the standard warranty provided and in case of major issues, replaced completely.

The finalized update regarding these will only be available by March 2019. The indicative image of the product is given below

Stay tuned for the updates ! And thank you for the support 🙂

Unexpected delay: Project RDOne

Dear Well-wishers,

We are sad to announce that the timeline of the first project RDOne will be extended to April 2019 for the production release. This delay is due to audit, taxes and related administrative regulations of the company. We will, be rolling out the prototypes in a few weeks and will be tested for about 1 month before being brought out for the initial release, which is expected to be in April 2019, the new financial year beginning.

We are taking in time so as to make the prototype model as close to the final production model as it can be. We expect the basic model to be ready in a few days.

Regards,
The Team,
RegalDream Technologies

What do we build ?

Passion to learn, is not always limited to the theoretical things. For some, they find theory to be interesting, and yet a few, doing things practically makes them feel better. While there are pros and cons for both, doing a bit of both can be exciting. As far as I, the author is concerned, for me, it has been something like seasonal or changes frequently.

There were days when I loved doing theoretical things, and then minutes later, I might be a completely different maniac. Let us not get too personal !

To begin a company, has always taken a lot of though behind it. Since we didn’t look into starting a software start up, we always knew that we would need some serious investments, unlike most other start ups which focuses on software. Competing with the giants would be difficult. Yeah, it always is.

‘World is never fair’ Being a start up, entering in the competitive segment of electronics product manufacturing, with no much prior experience, doesn’t exempt the company from competing against the giants in the market. Unlike the software counterparts, the hardware start ups, had to clear up the product line before taking on the market. There is no over-the-air update for hardware ! Moreover, every time to change the design, it costs you money and time, a lot of time !

Still after knowing all these, we wanted to build. When we looked into the technical advancements that took place, and the floods in the electronics market, especially in the mobile segments, we knew that there were things that people needed. My handset from 2016, the Redmi Note Prime, has a similar performance and specification as the phone now, not same, but similar. 3 years! Gradually, people had attained a will to spend more. A phone, which works, but is 2 years old, is generally exchanged for a new, slightly improved version for a descent enough price. That change in the spending power, and the will to adapt new technologies, has lead to a bigger market where the start ups could survive.

We also understood what the people needed and what they would be willing to buy. And we eventually learnt to build smaller things. maybe not perfect. We might learn to do that eventually. People needed things that made their lives better, and which saved their time. That is what we will be making !

RDOne: Prototyping update !

We have our RDOne Electronics hardware prototyping almost in its final stage. We had earlier gone for white case colour in our 3D Print designs. We have changed that to Black and will mostly be matte finish.

We have, in this updated case design, accommodated indication on skin display indication. We have also rectified the design changes, and improved the placement with legs for the case. This design will be almost ready for production, and we hope it will be 🙂

Our version 5 of case design before being processed. Image taken just after being 3D Printed.

Stay tuned for more updates. We will, hopefully, be getting back with the prototype in a week or two !