SlideShare verwendet Cookies, um die Funktionalität und Leistungsfähigkeit der Webseite zu verbessern und Ihnen relevante Werbung bereitzustellen. Wenn Sie diese Webseite weiter besuchen, erklären Sie sich mit der Verwendung von Cookies auf dieser Seite einverstanden. Lesen Sie bitte unsere Nutzervereinbarung und die Datenschutzrichtlinie.
SlideShare verwendet Cookies, um die Funktionalität und Leistungsfähigkeit der Webseite zu verbessern und Ihnen relevante Werbung bereitzustellen. Wenn Sie diese Webseite weiter besuchen, erklären Sie sich mit der Verwendung von Cookies auf dieser Seite einverstanden. Lesen Sie bitte unsere unsere Datenschutzrichtlinie und die Nutzervereinbarung.
March 6, 2015
• HAX (aka HAXLR8R) invests in hardware startups and makes sure they build the right
thing, build it right, and get to market fast.
• We see well over a thousand hardware startups per year across categories ranging from
robotics, sensors, health tech, smart home and lesser known ones such as sports tech,
pet tech, bio-sensors and more.
• As a result, it gives us a sense of the “near future” - products that might launch in a
year, later or will never get sold. We try to play our part in bringing the most promising
ones to market.
• This report has been created to give an overview of the “State of Hardware”: innovative
products but also aspects of prototyping and manufacturing that often go unreported on
the arduous journey to success or oblivion.
• Comments are welcome on how to improve this report at email@example.com or via
twitter at @haxlr8r.
• Applications to the accelerator program are at www.haxlr8r.com.
Benjamin Joffe, Cyril Ebersweiler, Duncan Turner
Shenzhen, March 2015
8. AR / VR
11. Twelve wares to avoid
14. China Rising
1. Hardware Trends
2. Fundings & Exits
3. Ecosystem Growth
5. Personal Health
6. 3D Printing
7. Smart Home
Table of content
1. HARDWARE TRENDS
Source: Aerial Screw by Leonardo da Vinci
Smart watches and trackers, augmented reality,
smart home devices, robots and self-driving cars…
technology is in the news and on shopping lists.
Hardware startups are on the rise across
existing and new categories.
Will every object become ‘sentient’ and connected?
Will every object become ‘intelligent’ and
autonomous thanks to onboard processing?
• Falling prices and advances in
computing, sensors, batteries and
connectivity have ushered a wave of
• Those connected devices (smart
watches, trackers, sensors…) largely
rely on the computing power of
smart phones, or the cloud.
Sentient and connected
Some “ﬁrst wave” devices
with sensors and connectivity
Bluetooth, WiFi, Zigbee and cellular
New M2M solutions
Cellular dev kit
with data plan
Cellular network for IOT
$214k on Kickstarter
as of Mar 6 ,2015
Raised $4.9MRaised $148.4M
• Differentiating and maintaining a
competitive advantage is harder than ever
as components are getting commoditized
and products are global from day one.
• As a result, a ﬂood of smart watches, activity
trackers, ﬁlament-based 3D printers and toy
drones are entering the market.
• The way out might be in new sensor
technologies (non-invasive or embedded),
design, software, AI and communities of
users or developers.
• New applications in sports, preventive and
personal health are creating emerging
behaviors toward human augmentation.
Sitting & ECG tracker
Mini-quadcopter for $14
on China’s Taobao
Gloves are off! Who was ﬁrst? Who’s winning?
Gloves are off! Who was ﬁrst? Who’s winning?
Gloves are off! Who was ﬁrst? Who’s winning?
• Prototyping is easier and cheaper
thanks to various platforms such as
Arduino, Raspberry Pi, Spark, 3D
printing and more.
• Time To Market (TTM) is shortening
as the loop is closing between
manufacturing, logistics and retail.
• “Online-to-Ofﬂine” (O2O) is rising:
complement an online store with
ofﬂine fronts or pop-up stores with
low or no inventory.
Faster to market
Arduino, Raspberry Pi
and Spark help with prototyping
help launch new products
“Rent a store front”
• With smaller and cheaper MCUs,
objects are turning into computers
able to process sensor data and run
• Next step might be “apps on things”
and have truly “enchanted objects”.
Toward enchanted objects?
The power of an iPhone
on your wall
On-board face recognition
Nest Thermostat Welcome
IP camera by Netatmo
This camera uses STAK
technology to run apps
• Today, low-cost automation, 3D printers and
robots are expanding to new industries and
entering workshops, labs and homes.
• Drones have found applications in
entertainment, imagery, surveying and exploring
• Desktop 3D printers have expanded to include
materials such as metal, carbon ﬁber, glass and
organic materials such as chocolate, skin and
Beyond consumer devices
Will we train sports
Print faster and better
Drones for fun, imagery and delivery
Search & rescue robot
Hardware hype cycle:
Which technologies will get adopted at scale?
More hardware startups are getting funded,
supported by the success stories
of recent acquisitions.
Consumer appeal and venture capital are not always
correlated: while consumers might not care about
those, defensibility, long-term strategy and unique
positioning are strong contributors to successful
Funding and exits
• In 2014, hardware unicorns were on the rise.
• The year saw billion-dollar acquisitions of Nest,
Oculus, Beats. GoPro went IPO.
• Several other recent M&A such as Dropcam,
Boston Dynamics, SmartThings, Basis reached
hundreds of millions.
• In a mere 4 years, the Chinese smartphone maker
Xiaomi went from zero to the largest market share
in China. Its valuation is now over $40 billion,
diversifying into more connected products.
• Over 200 hardware investment deals ranging from
seed to mega-rounds. Many more unannounced as
investors are tip-toeing into the hardware ﬁeld.
Zero to One... billion dollars
Hardware startups on AngelList
Source: AngelList, 2015
• As of March 2015, there were 3,022 hardware startups
Hardware startup investment
Source: Tomasz Tunguz, Redpoint Ventures with Crunchbase data, 2014
• There is a notable increase of investment in hardware
startups since 2013, in particular for seed and series A.
Top hardware startups funding
Note: VC funding amount was found for 38 out of the 41 top hardware companies.
Data source: Crunchbase, Jul 2014
Top hardware startups valuations
Note: Known valuations of 20 largest hardware startups, Jul 2014
• Overall VCs fund mostly (1) Serial entrepreneurs (2)
Demos with "Wow!" (3) Growth.
• This creates a funding gap (“Bridge of Death”)
between a demo and growth. As a bridge, startups
use crowdfunding, grants, pre-sales,… and get to
• Most startups plan only their ﬁrst product, which
reduces their chances of funding.
• Other stumbling blocks: manufacturing, ﬁnancing,
VCs warming up to hardware... to a point
Cash position over time
PROTOTYPING PRODUCTION SCALING
Bridge of Death
• Since the beginning of the crowdfunding
platform Kickstarter, $389M have been pledged
across 5,500 technology projects.
• Beyond direct sales, crowdfunding is an
“awareness enabler” for distributors, investors,
developers and supporters.
• Several hardware companies were “born” from
crowdfunding: Oculus, SmartThings, Pebble,
• Technologies aiming for B2B applications can
also start with a consumer product to increase
• Yet, manufacturing remains a barrier. Most
unprepared projects ship late, if at all.
Crowdfunding is an enabler
• Very few campaigns (if any) are an “overnight success”.
• Platform-originated backers are a minority. Creators
have to generate the attention via media contacts and
community building. Media momentum is hard to achieve
yet is key to attract customers. Media have their own
schedule, which is rarely yours.
• Top campaigns often raised venture funding prior to
crowdfunding and spent on advertising, PR, or took a long
time to build a meaningful mailing list. Some had celebrity
• Credibility and genuine enthusiasm seem to matter more
than “video quality”.
Crowdfunding success is rarely an accident
• At the time of writing the Coolest Cooler,
a smart cooler for outdoor parties, was
the largest Kickstarter project with over
$13M in backing (go Pebble!). It rose
from the ashes of a failed ﬁrst
campaign, with better preparation and
timing (it ﬁrst failed a campaign run
during the Winter).
• The early media coverage of the
smartphone printer Prynt by TechCrunch
got shared over 60,000 times, generating
more than 400,000 views of a casual
demo. This helped build a waiting list of
over 50,000 people. Prynt raised over
$1.5M a few weeks later.
Anatomy of a campaign:
Two successful projects
Kickstarter top 10 hardware projects
$13.3M $10.3M $8.6M $6.2M $3.4M
$3.4M $2.9M $2.8M $2.4M $2.4M
Failed its ﬁrst
Sold 10,000 units
of another watch
and raised $375k
DASH FORM 1
Indiegogo top 10 hardware projects
$2.3M $2.2M $2M $2M $1.7M
$1.5M $1.4M $1.3M $1.3M $1.3M
Feasibility in question
GEEK WAVE AIR TAME TRACKR
• Most crowdfunded projects are not suitable for
venture capital. Only a quarter of projects above $100k
raise VC money.
• There is some correlation between backing amount and
VC funding. Could one cause the other? Which one?
• Eventually, long-term success do not seem to be
correlated with crowdfunding amount so far.
Crowdfunding to VC funding
Source: Flybridge Capital Partners,Aug 2014
• Quirky’s community of inventors and
designers pitch ideas online, or help reﬁne
other ideas. The selected projects get built
by Quirky’s product design staff and sold
online and in retail.
• Quirky pays back 10% of product sales, split
between the initial inventor and other
contributors. So far, most projects are fairly
low-tech and best-sellers dominate sales.
• Quirky raised $185.3M in venture funding. In
February 2014, it had close to 300
employees, sales reached $100M in 2014.
The Quirky approach
The Maker Movement has gathered
considerable steam over the past few years.
How many makers will make the leap
to become a hardware startup?
• More makers, more startups, more events, more
funding. Everything is growing.
• Hardware is also getting more attention from both
media and investors, notably thanks to visible
success stories such as GoPro, Nest and Fitbit,
and to the rise of crowdfunding platforms.
• Yet, the ecosystem is not growing evenly in
terms of geographic distribution, availability of
tools, support, talent, capital and manufacturing
capabilities. Some places are better served than
others, and several retain strategic advantages.
A bigger ecosystem
• More creators are jumping into hardware thanks to
lower barriers of entry.
• Hackerspaces, TechShops, Fab Labs and various
incubators, public or private like France’s Usine.io
offer places for them to work, use tools, learn and meet
other creators. They often support the early prototyping
stages and act as “pre-accelerators”. Companies like
Wearable World also help projects get attention from
media, investors and brands.
• There are hundreds of Maker Faires, large hardware-
related meetups (the ones in SF, NYC, Waterloo,
London and Paris have thousands of members each),
thousands of Open Source Hardware projects, and a
growing number of events related to hardware and IOT.
• Platforms like Upverter, SupplyBetter and Hackster.io
help source manufacturing partners for later stages.
Early stage support for hardware
is getting more widespread.
Over a thousand hackerspaces
are active worldwide
37Source: http://hackerspaces.org/, March 2015
Over a thousand hackerspaces
are active worldwide
38Source: Renee DiResta, OATV, 2014
Hundreds of Maker Faires
are held worldwide every year
39Source: MakerFaire.com, March 2015
Meetup community growth
40Source: meetup.com, March 2015
Hardware Meetup Groups
IOT Meetup Groups
• The number of meetups and
their membership are growing
• Close to 20 groups have over
1,000 members. Events
routinely gather hundreds.
• The most active locations are
San Francisco, New York,
London, Bangalore and Paris.
• A strong second group is
composed of Barcelona, Tel
Aviv, Stockholm and Austin (TX),
Reston (VA), Washington (DC).
Popular IOT and Hardware meetups
Source: meetup.com, March 2015
41Source: meetup.com, March 2015
# Meetup Location Members
1 IOT London UK 4,712
2 SF HW Startup USA 4,086
3 IOT Bangalore INDIA 3,219
4 IOT SF/SV USA 3,074
5 Hardwired NYC USA 3,028
6 SF IOT USA 2,841
7 IOT Paris FRANCE 2,397
8 NYC HW Startup USA 2,319
9 IOT Central NYC USA 2,013
10 NOVA Makers (Reston, VA) USA 1,979
11 IOT Israel ISRAEL 1,776
12 Sensored (SF) USA 1,680
13 IOT Barcelona SPAIN 1,594
14 HacDC (Washington, DC) USA 1,342
15 IOT Stockholm SWEDEN 1,276
16 SF Wearables USA 1,064
17 Austin HW Startup USA 1,052
18 HW Startup Lab (London) UK 1,046
• At the end of 2014, there were over 2,000
startup accelerators worldwide. Their
structures vary: investment, corporate,
sponsored, non-proﬁt… with different
degrees of alignment with startups.
• Most focus on software. As a result,
hardware startups are often isolated and
can’t get the guidances and tools they need
to prototype and build products at scale.
• Hardware startups increase their chances
by connecting with suitable ecosystems as
well as building manufacturing and supply
Most incubators and accelerators
can’t answer the needs of hardware startups
Living next to an electronics market
will speed up prototyping
Makers and startups
Makers Build for fun, education, goodwill, etc…
Turn their hobby into a business.
Often create tools for other (merry) makers.
Invent and sometimes license their ideas.
Born to scale.
Focus on proof-of-concept.
Focus on manufacturability
and supply chain.
Reduce bill of materials, care about
component availability and life cycle,
integrate supply chain.
Focus on logistics, distribution
and cash ﬂow.
Find ways to ﬁnance inventory,
protect margins and scale up.
Required skills of hardware startup founders
Things get real with prototyping
An object representing the ﬁnal product. Does not work.
Manufacturability or cost are often not considered.
Proof of Concept A device performing - to some extent - the intended functions.
A prototype that works.
Size, design, cost and performance are secondary concerns.
Works, with a design close to what the ﬁnal product.
Works, with design, manufacturability and costs carefully
considered. It is more or less identical to the ﬁnal product.
One of few units coming out of the assembly line prior
to full production.
From idea to product:
Leap Motion Controller
• “Hardware is hard”.
But what is hard exactly in hardware?
Once the R&D part is covered, the riskiest parts are
often in reducing costs to make the product viable,
and handling manufacturing.
• Crowdfunding backers typically invest in early
prototypes (when not mere renders or form factors).
Those might not have completed the critical R&D and
• Backers, media and investors are often wowed by
demos and underestimate the difﬁculties of both
manufacturability and manufacturing.
And that’s when a product can be made at all!
• All would beneﬁt from a better understanding of the
milestones the creators have cleared, so as to grasp
both the level of risk and the level of support needed.
Risk in hardware startups:
Does it work? Can it be made? Can it scale?
Some parts can be
hard to source
Source: Pokeball, Pokemon animation series
Many daily objects are getting ﬁtted
with sensors and connectivity
Can the market sustain the many
smart watches and trackers?
Eventually, “wearable” is not a category,
what matters is the problem the device is solving.
• Over 40 companies offer Android Wear
smart watches. Samsung, Motorola,
Sony, LG, Asus and others shipped
an estimated 720,000 units in 2014
(source: Canalys, 2015.2).
• 7-years old startup Pebble cut prices in
2014 and reached over 1M units since
its introduction. It added Android Wear
app compatibility, opening up its
ecosystem but eroding differentiation.
• Initial orders for the Apple Watch due to
launch in April 2015 are estimated
between 5 and 6 million.
Will the market be big enough for all
the smart watch makers?
Differentiation is becoming difﬁcult
First generation trackers disappoint
Will the next generation fare better?
gave up on the
Will the next
stick and be
Evolution of trackers:
Toward fashionable or invisible devices
Raised $66M Raised $520M Team reportedly ﬁred
China’s Xiaomi launches a $13 activity tracker
Mi Band by Xiaomi
Yours for $13 (mostly in China so far)
• Step count
• Calories burned
• Sleep tracking
• Unlock phone
• Incoming call alerts
• 30 days battery life
• Water resistant
Multi-purpose and focused trackers
Golf / Baseball / Tennis
by Shot Stats
Surf / Snow / Skate
Raised $15M $120k on Kickstarter$103k on Kickstarter
In many places, there are more pets than babies!
Dog activity monitor
$251k on KickstarterRaised $240k
global location tag
(using cellular triangulation)
$145k on Kickstarter
as of Feb 22, 2015
$2.6M in crowdfunding
From learning to daily practice
Midi smart guitar
Automatic guitar tuner
$353k on Kickstarter $178k on Kickstarter#HAX
5. PERSONAL HEALTH
Source: Mr Spock with medical Tricorder, Star Trek TV series
New low-cost and non-invasive sensors are enabling
a new wave of personal health devices.
Personal health devices
Scout by Scanadu
ECG, breathing, Temp
Air quality monitor
Sitting & ECG tracker
Wink by Kindara
$2.8M on Kickstarter
$1.7M on Indiegogo
$225k on Kickstarter Raised $1.6M
June by Netatmo
3D printing has expanded from basic prototyping
with plastic to numerous materials.
Future printers might produce commercial-grade
products and allow for micro-manufacturing.
Toward more mature technologies
• 3D printing basics:
• FDM (Fused Deposition Modeling) uses
heat to extrude a ﬁlament of plastic material.
• SLA (Stereolithography) uses a beam of
light to curate a photo-reactive resin.
• New developments:
• Cheaper printers, portable printing pens.
• New desktop technologies.
• New materials, multi-material prints, multi-
color prints, even printing houses!
• Toward factories with 3D printing farms?
PLA (polylactic acid)
SLA cures polymer
Acq. by Stratasys for $403M
FDM: the race to the bottom
$3.4M on Kickstarter
$1.4M on Kickstarter
$1.6M on Kickstarter
$599 on KS
$457k on Kickstarter
£255k on Kickstarter
First with cool ink
$2.3M + $1.5M
First 3D printing pen
$205k on Indiegogo£732 on Kickstarter
FDM: printing pens
$99.99 $139.95 pre-order $119 pre-order 04.2015?
$2.9M on Kickstarter
$820k on KickstarterOpen source
SLA: the new frontier for desktops?
$3,299 $5,995 pre-order $2,999 pre-order
SLS (Laser sintering)
$213k on Indiegogo
Prints a variety of materials ranging
from plastics to ceramic or metals
Launching in 2015
10x faster than classic FDM, with higher quality,
allowing production runs
Metal Carbon ﬁber
Bio ink Skin & Bone
Custom 3D printed ear buds
$767k on Kickstarter#HAX
Custom 3D printed in-soles
A giant 3D printer builds 10 houses in one day
Access, indoor comfort, smarter appliances…
the house is getting connected.
Will anyone win the battle for the home hub?
Will things be interoperable?
Google is making moves toward owning home data;
the market is waiting for Apple’s move.
The smarter home
Acquired by Google for $3.2B
Acquired by Samsung for $200MAcquired by Google
WiFi IP camera
Acquired by Google
Raised $5.8M Raised $700k#HAX
Designed byYves Behar
Raised $2.2M via crowdfunding
Smart body analyzer
$503k on Indiegogo
$2.4M on Kickstarter
Smart bed cover
$936k on Indiegogo
(as of March 6, 2015)#HAX
Home sensor networks
Most augmented and virtual reality products
are not commercialized yet.
Will Christmas 2015 be their coming of age?
Will 2015 be the year of AR/VR?
• The most iconic AR project has just
been discontinued: Google decided
to stop the production of Glass less
than two years after launch.
• Glass found a number of niche
applications but faced severe
criticism regarding privacy and failed
to reach mass market adoption,
partly due to its high price tag of
2013.04: Glass is introduced to “Explorers”
2014.05: Glass open to the general public
2015.01: Google stops producing Glass
Hired Neil Stephenson
Sci-Fi author of “Snow Crash”
Notable crowdfunded projects
$194k on Kickstarter
Raised $23M from VC
$1M on Kickstarter
$2.4M on IndieGoGo
Hired Steve Mann,
pioneer of wearable tech
as Chief Scientist
Founded by former
Valve Software employees
OSVR by Razer
Open Source VR
$2.4M on Kickstarter
$2B acq. by Facebook
by HTC and Valve
Includes controller and laser for sensors
Samsung Gear VR
Powered by Oculus
Other notable projects
$261K on Kickstarter
iPhone VR headset
patented by Apple
• Giroptic’s camera will be the ﬁrst 360 degree video
supported by YouTube.
• The availability of such content will make virtual
reality an increasingly attractive proposition.
Solving the creation & distribution of 360 video
• Most virtual reality experiences are limited to display.
Interfaces like keyboard or mouse are not convenient.
• The Leap Motion controller can be combined with the
Oculus Rift to bring a user’s hands into the virtual
Leap Motion solving the “hands” problem in VR
Capturing the world in 3D
$1.3M on Kickstarter
Source: Viper probe droid of Star Wars Episode V: The Empire Strikes Back, Lucaslifm
Drones have found applications
in entertainment, imagery and surveying.
E-commerce giants Amazon and Alibaba and others
are looking into using drones for deliveries.
Piloting, handling obstacles as well as
autonomous ﬂight remain challenging.
Regulations are slowly catching up.
The state of drones
Some drone applications
Enterprise / Surveying
Imagery / UAV
Cirque du Soleil
#HAX Raised $85M
Raised $19.7MMarket cap: $247M (March 2015)
Amazon and Alibaba
Successful trial in Feb 2014
US regulations don’t allow
deliveries by drone so far
Tea packages were delivered
to areas close to distribution centers in
Beijing, Shanghai and Guangzhou
FAA proposed rules that include a knowledge
test, registering the drone and stay under 500
feet and within line of sight.
Source: Techcrunch, Feb 2015#HAX
Collision-happy inspection robot
developed at EPFL
Distributed Flight Array (DFA)
Self-assembling ﬂying robot
developed at ETH
Developed at ETH
Developed at ETH
Source: JARVIS AI system in Iron Man, Marvel Movies
A robot can be described as
“A machine performing complex actions
in the physical world”.
Most don’t look like humans but enjoy
increasing levels of autonomy and intelligence.
Today, low-cost and smart robots
are expanding to new industries
and entering workshops, labs and homes.
Robots are coming
Expensive, simplistic or ﬁctional
Low-cost robots for ofﬁce, workshop and lab
Robot cooks, butlers and waiters
Pengheng Space Capsule Hotel
“Butlr” butler robot
Aloft Cupertino Hotel
by Momentum Machines
Guards and sales assistants
OSHbot by Lowes
Sales assistant robot
Security guard robots
Raised $6.7M from VC#HAX
Cleaning and painting
Autonomous painting booth
Autonomous cleaners & lawn mowers…
First version sold in 2002
>10M units sold since start
First robot sold in 1995!
…are getting commoditized quickly
They can now be
China for less
than a quarter of
the US retail price.
getting tough for
simple robots like
and lawn mowers
as they go
Nao and Pepper
by Aldebaran Robotics
Raised $30.7MAcq. by Softbank, $100M
by Double Robotics
Toys and drones
Robotic mouse for cats
Sphero & OllieVarious drones
From research labs to schools and homes
Lego Mindstorms Makeblock
• ROBI is a robot kit created by Tomotaka
Takahashi from ROBO-GARAGE. It is
sold via a weekly magazine published by
• Readers receive a few parts every week
with detailed information. 70 issues are
needed to build the robot ($20/issue,
• It sold an estimated $100M by Jan 2015.
ROBI: DIY subscription robot
From surgery to soft robotics
for mother/infant communication
for premature babies
• Google acquired several companies involved in robotics,
vision and control.
• Several of the projects were ﬁnanced by DARPA.
Google goes robotics
by Boston Dynamics
by Boston Dynamics
Underwater and surface robots
Underwater Exploration Robot
Oil spill cleaning robot
$112k on Kickstarter#HAX
by Empire Robotics
Other robotics novelties
“You’re just as good as your grippers”
DIY ping-pong robot
Cycling robot by Masahiko Yamaguchi
Robots for factories and warehouses
Versatile factory robot
by Rethink Robotics
Acquired by Amazon for $775M
15,000 robots are in operation
across Amazon’s 50 US facilities
Source: CNET, Nov 2014#HAX
Over 40 unmanned trucks are
operated by Australia’s mining
giant Rio Tinto.
Each loaded truck weights
over 500 metric tons.
Source: Mining.com, Sep 2013
Not exactly robots but still interesting:
Personal mobility solutions
• The market was an estimated $800 million in 2013
and might grow 20x to over $16 billion by 2020
(Source: WinterGreen Research, January 2014).
• According to The Robot Report (July 2014):
+ Robotic harvesting, irrigation, pruning, weeding and thinning
devices are being ﬁeld-tested all around the world.
+ Robotic spraying and seeding have been going on in Japan
and Australia for years.
+ Driverless tractors are getting started.
+ Robotic cow milking is growing.
+ Nurseries are beginning to use pick-and-place robots.
+ Aerial observation robots might support agricultural precision.”
Crops, cows and calves
robotic cow milking
11. THE 12 “WARES” TO AVOID
There are many ways to fail at a hardware project.
Getting the wrong market, timing or positioning
are enough to wreck a startup.
Avoid the following twelve “wares”.
Recognizing good hardware startups
"Happy families are all alike;
every unhappy family
is unhappy in its own way.”
Leo Tolstoy, Anna Karenina
Anna Karenina Principle
Beware of those 12 “wares”
Solution looking for a problem
Can’t be made
Compromised beyond reason
Building the wrong thing
No market yet
Too tied to local conditions
• Too small market.
No chance of a larger one.
• Includes FUNware and
FUNware: Rubik’s cube solving robot
Guinness world record holder
ARTware: One of Japan’s teamLab
outstanding tech/art installations
NICHEware: This robotic trash can
will catch some of your throws#HAX
• Too easy to copy.
• Trivial engineering and
market demand will
attract competition and
• Defensible intellectual
property is not limited to
patents: it can be
software, trade secrets or
a community (such as
Makerbot and GoPro).
Tile keeps track of your things with bluetooth
It now has droves of competitors
Smart thermometer on Indiegogo#HAX
• Lack of differentiation.
• A weak positioning will
lead to limited sales,
even after an initial
launch via crowdfunding.
• Your “better mousetrap”
needs to be multiple
times better in some way
(price, speed, usability…)
than existing solutions to
capture market share.
Over 50 companies launched 3D printers
using crowdfunding and raised over $100k
Source: Flybridge Capital Partners, 2014.06#HAX
• A solution looking for a
problem (“a hammer
looking for nails”).
• Academic research often
falls into this category.
“Cubes could transform into a chair or a desk”
Source: MIT News, Oct 2013
“They’re trying to get it in the hands
of engineers with big ideas”
Source: Engadget, Oct 2014
“A glimpse of a future
[…] that still feels far away”
Source: TheVerge, Jul 2014
Light ﬁeld photography camera.
Now shifting from cameras to virtual reality.
Self-assembling robots from MIT
$510k on Kickstarter
• Can’t be made.
• Includes NAIVEware
$1.6M on Indiegogo
“A clear cut example of a few guys
with a neat idea grossly underestimating
what it takes to develop a product.”
Source: DropKicker, Aug 2014
“Tony Hawk has now issued a video apology
for his role in the HUVr Board prank”
Source: Heavy.com, Mar 2014
“Never attribute to malice that which is
adequately explained by stupidity”
Ada by Triggertrap
for high speed photography
£290,386 on Kickstarter
The creator posted “How our
$500k Kickstarter campaign
crashed and burned”, canceling
the project and refunding the
remainder of the money to
“Backers found problems with HW & SW,
including a lack of several advertised features.”
Source: PC World, Sep 2014
“The bulky size [and the need] to have
the app open negates any time-saving”
Source: Snazzy Labs, Nov 2014
Kreyos smart watch
raised $1.5M on Indiegogo
Ring bluetooth controller by Logbar
Raised $880k on Kickstarter.
GoBe calorie counter
Raised $1.1M on Indiegogo
“It does not deliver on its most exciting feature”
Source: Engadget, Feb 2015
• Successfully building
the wrong thing.
• This happens to small
and large companies
• Bigger companies might
ship bigger failures.
Microsoft Tablet PC (2002)
“The tablets we had done weren’t as thin,
they weren’t as attractive.”
Source: Bill Gates, Jul 2012#HAX
• Validated a market but
woke up competitors.
• Delays mean sales start
later, putting pressure on
• Delays also mean
competitors might enter
• In some cases, delays
lead to obsolescence.
Their 14,000+ backers waited over 2 years.
Meanwhile, competitors entered the market.
$2.2M in crowdfunding.
• Minimal or negative
• Healthy margins allow to
cover the bill of material,
tooling, returns, salaries and
promotion costs. Retail also
takes a considerable share.
• Some products might use a
different business model
allowing them to offer
hardware for cheap or for
free (maybe even pay
“You will be receiving Shru at cost price.”
Creators will have to ship
over 4,500 units at cost.
Electronic cat toy
Raised $171k on Kickstarter
• People stop using them.
• Category pioneers often
have ﬂaws. Later versions
might overcome them
and help grow a market.
• For wearables, the next
generation of devices
might ﬁx some key issues
like battery life and live
feedback and trigger
mass market adoption.
• Ahead of its time.
• Being too innovative can
be a death sentence.
Got discontinued within 2 years of launch.
Usability, lack of “killer apps”, price
and social barriers prevented it
from reaching the mass market.
Nabaztag (Violet, 2005)
Aibo (Sony, 1999)
• Too tied to local
• Peculiarities of local
ecosystems can prevent
Japan’s ﬂip phones still represented
over a quarter of all shipments in 2014.
For a decade, those phones have had
very advanced functionalities including
apps, mobile TV and NFC payment.
The isolated technological path
followed by Japan is now often called
Phone with electric shaver
Part of the shanzhai “mass production artwork”
production system in China
Safety regulations, IP and logistics will prevent
exporting to most countries.
Prototyping has become dramatically faster
and cheaper for electronic products.
How far are we from building hardware
at software speed?
• Barriers for prototyping are falling.
• Mechanical: 3d printing, laser cutting,
CNC machining, vacuum forming…
• Electrical/SW: some prototyping
boards are now production-grade,
circuit printers coming to market.
• Electronics: prices falling,
• Robotics: DIY / open source kits.
• Connectivity: chips, modules,
• After early prototyping, using 3D printers
quickly becomes a time sink.
• In addition, most additive techniques
can’t be used in manufacturing.
• So once form factor is clariﬁed, use real
CAD data and move away from 3D printing.
Outsource early in order to polish
• Leverage factory expertise and blend
prototyping, DFM and manufacturing,
toward making a real product.
• Leverage the supply chain and take into
account component availability and life-
3D printers usage stops at proof of concept
Time spent ﬁxing a printer
or re-doing prints can be
better spent elsewhere
• Lots of mechanical parts can be found off-the-shelf or through kits.
• If in China, get any custom part in a few days, including motors.
Worldwide delivery can also be arranged for many parts, including
• Below, the parts included in the Makeblock robotic kit allows the
building of fully functional machines such as a 3D printer or a laser
DIY, Open Source, etc.
3D printer Plotter / CNC / Laser cutter
An OS built for
Wi-Fi and Cellular
• The spread of prototyping and educational platforms contribute
to the growing number of hardware startups.
A slice of Pi for everyone:
5M units sold in 3 years since launch
2012 2013 2014 2015
3D printers usage stops at proof of conceptZero to ﬁnal prototype in 3 months
Prototyping speed is accelerating:
what used to take a year
can be done in a few months.
Source: Modern Times, Charles Chaplin
Manufacturing is often where hardware startups fail.
Integrating the supply chain early in the development
process can dramatically increase odds of success.
“Get out of the building”
- Lean startup principle
“Get on the factory ﬂoor”
- Lean hardware principle
Lean Hardware Startup
• To start: de-risk their supply chain to ensure supply of all parts.
• Further: own relationships to be ﬁrst-class customers and even block
competitors from sourcing the best parts.
• Eventually: own their assembly (or control it like Apple with Foxconn):
control processes and machines.
What every hardware startup wants…
“Apple has exclusive deals with hardware
manufacturers for the best parts for iPad.
[…] HP could only source the second best.”
Source: TheNextWeb, Aug 2011
Can’t touch this
“Apple Bought $578M Worth Of
Sapphire In Advance.”
Source: TechCrunch, Nov 2013
• Startups are not like Apple: volumes, cash, inﬂuence,
(very) long runway.
• “Apple quality” takes time and is more pricey
(machined aluminum? laser-made holes? etc.).
• Seeking “perfection” can cause important delays and
…and can’t have
• Creating a product that has never been made
before is a difﬁcult task.
• Design with the factory: avoid mistakes
thanks to “Design from manufacturing”.
• Manufacture in the right location
with access to the relevant supply chain.
• Build with the right partner: toy factories
are great to make toys, less so for robots.
• Launch fast and launch early: improve your
product and supply chain during and
between runs, on the factory ﬂoor by
manufacturing in small batches.
Leveling the playing ﬁeld:
Partnering with factories
• Without hardware and manufacturing expertise a
company is at risk of becoming a “hollow company”,
unable to plan or discover improvements.
• Successful design and manufacturing requires
knowledge of the tools.
• Manufacturing issues are hard to solve at a distance or
with middlemen: improvements and discoveries can also
be made on the factory ﬂoor.
• Using middlemen makes it harder to adjust your supply
Startups need in-house experience and
“Get on the factory ﬂoor”
“The problem is that we don’t understand the problem”
- a hardware startup.
Solving problems is easier on site
• Mastering the supply chain
requires ﬁnding the original
source for everything: is this
supplier really THE supplier?
• "Parachute manufacturing” -
e.g. a week overseas to source
a contract manufacturer - is
rarely enough due diligence to
select a reliable long-term
• The depth of the supply chain,
combined with knowledge of
manufacturing and materials are
long-term strategic advantages
against rapid commoditization.
Supply chain due diligence and management
“The biggest roadblock to the success
of hardware startups isn’t money,
machines, or material: it’s ﬁnding
the right partners and people
to implement their vision.
Would you hire an agent to shop
for dinner or buy clothes for you? […]
After all, “It’s people! - supply chains are
made out of people!”
- Andrew “bunnie” Huang,
bunnie studios blog, Dec 2014
• PCB assembly: the vast majority is now fully automated.
• Injection molding & CNC machining: operated with little
overhead, high throughput and robot arms.
The dream of a fully automated assembly line
is getting closer by the day.
The automation equation
…and 1 million robots…Less employees… … typing on a touchscreen.#HAX
• Prototyping will remain easier when located next
to strong component supply chain.
• Some automated small batch manufacturing
(total <1k units) could gain a local cost and time
advantage thanks to high-quality prototyping tools
used as “printing farms”.
• Product assembly might become more distributed
to facilitate shipping and on-time deliveries. Human
and robotic assembly lines will get closer to their
One future: distributed manufacturing?
China used to be known for only making
cheap copies at scale.
Today’s china also makes the world’s
highest quality products.
The beneﬁts of China’s supply chain for speed, costs
and scaling from a prototype to millions of units
is now open to startups.
A tale of two Chinas
• Copying China’s electronics
supply chain would be as hard
as copying Silicon Valley’s
• From rare earth production used
in electronics (China produces at
least 70% of the world’s rare earth
elements) to electronics
manufacturing, assembly and
supply chain, it is unlikely to
move anytime soon.
• Instead, combine China’s supply
chain & manufacturing know-how
with global market access. China
is also a huge consumer market
for electronics and building things
there helps understand that.
The hardware world is not ﬂat
• Shenzhen, the world capital for electronics and
supply chain is now seen as the “Silicon Valley for
• Shenzhen’s ecosystem is also used to prototype
better, faster and cheaper: design with local
components and take advantage of the 24-hours
Does all hardware lead to Shenzhen?
“The city has a complete ecosystem of low-cost labor, massive
factories and leading manufacturing technologies, making it
able to turn out almost any kind of hardware on a large scale
Both Shenzhen and Silicon Valley have a critical mass. We’re
most likely to be successful connecting with Shenzhen than
competing with it head on.”
“A week in Shenzhen is worth a month in the Valley.”
- a hardware startup founder
Connecting with Shenzhen
MIT Media Lab
Shenzhen map for makers
Source: Huaqiangbei map for makers
• The electronics market is
made of over a dozen
multi-story buildings ﬁlled
with shops selling
components of all kinds.
• Most shops are tied to
factories and can supply
from 1 to thousands of
Shenzhen electronics market
Source: Dangerous Prototypes
SEG electronics market Shops in the SEG electronics market
Magnets of various sizes Soldering workshop in mobile repair shop
Top universities building closer ties
• Several leading research
institutions are building
ties with Shenzhen’s
• Notably, Berkeley, MIT’s
Media Lab and Center for
Bits and Atoms have
taken steps in that
• The global Fab Lab
conference Fab 12 will be
held in Shenzhen in 2016.
• Barriers to get to market have gone down
(with crowdfunding, e-commerce) and
OEMs are trying to go up the value chain.
• What they lack is design capabilities,
brand recognition, sales & marketing
and customer service.
• Design capabilities: some are hiring
designers and acquiring global talent.
• Brands: some are buying well-known
or distressed brands to get credibility,
distribution and intellectual property.
• Investment: some are investing in
other companies, including startups.
Rise of the OEM
The animatronic pet PLEO was
acquired by its Chinese OEM,
JETTA after ﬁling for bankruptcy
Famed designer Yves Behar
sold a 75% stake in his ﬁrm
Fuseproject to the Chinese
BlueFocus for $46.7M
• Several “creators” simply rebrand or repackage
products sourced on Alibaba.
• Some are visible on crowdfunding platforms.
“ALIware” as a new trend?
• Smartphone giants like Apple and
Samsung aside, very few foreign
hardware startups have
performed well in China.
• Launching in China requires to
adapt to an entirely different
• This includes different marketing
practices, distribution channels
and sometimes even a different
Foreign hardware startups in China
Misﬁt received a $40M
investment from Xiaomi and
other investors. China is now
the largest market for its Shine
OUYA, the android-based
game console received a $10M
investment from Alibaba.#HAX
• The “Pressy” smart button (inserted in the audio plug of a smartphone to
create a shortcut for services) was copied 3 times in China with
successive price drops until it became free with a new business model.
• The Internet company Qihoo gave away 1 million units to college
students to acquire users for its mobile services.
From Kickstarter to Free in 10 months
• Suffering from competition with large local players
(Xiaomi, Huawei, Lenovo, Midea, ZTE, TCL…).
• Extremely strong execution and iteration.
• Lack of effort in “disruptive” R&D.
Chinese hardware startups
$125M in revenue
• Launched by a team of veterans pursuing a platform/software distribution strategy rather
than pure hardware proﬁts. It sold 61 million phones in 2014, claiming China’s #1 spot.
• It sells inventory online directly to consumers as soon as it’s made, with no
advertising and relatively low margins. This approach allows for lower retail prices.
• Xiaomi leverages its community of users to guide product development. The company is
not a mere phone brand but a distribution channel to the aspiring middle class.
• In Sep 2013, Xiaomi hired Hugo Barra from Google to head its international expansion.
The Xiaomi case:
How it became China’s #1 smartphone maker
Founder of Xiaomi
Lei Jun Mi Note
A phablet for 368 USD
VP of International, Xiaomi#HAX
• Xiaomi is commoditizing numerous hardware products.
• It intends to launch 100 hardware products with OEMs and investments
(Misﬁt, iHealth, Yeelight etc.).
• It already offers smartphones, tablets, power banks, activity trackers,
headphones, TV box, 4K TV screen, smart home sensors & connectors,
webcam and more
Xiaomi expands beyond smartphones,
commoditizes more hardware devices
24 USD16 USD13 USD
Miband tracker Bluetooth speaker IP camera
4K TV screen
Xiaomi’s answer to GoPro? #HAX
• In March 2015 Xiaomi
launched a new action
camera built by its camera
• With its phones, Xiaomi
captured the middle market,
notably from Samsung and
some high-end from Apple.
• With its action camera,
Xiaomi might capture
market share from GoPro
and create a new market
Xiaomi offers many different mounts
• This report is for informational
purposes only and makes use
of various public and non-public
• HAX is an investor in several
startups mentioned in this report
• SOSventures is an investor in HAX
and in several startups mentioned
in this report
• HAX is a startup accelerator focused on hardware
4 months program in Shenzhen
Demo day in San Francisco.
• Most active investor in hardware
65 startups (B2B and B2C)
Robotics, IoT, sensors, smart home…
• Most experienced investor with crowdfunding
$300,000 average raise
• Pioneer of the “Lean Hardware” methodology
TechCrunch series & presentations
Apply to the next program:
3D printed drones
HAX robotics startups (1/2)
Ultra-fast 3D printer
Low-cost robotic arm
Software is from the Bay,
Hardware is from Shenzhen
Hardware: Harder, Better
8 things about crowdfunding
Why makers fail at retail
HAX at Stanford
Building lean hardware startups
Why Makers Fail At Retail
8 Things About
Breeding Hardware Unicorns
March 6, 2015