Sweet!

Making the interconnected future WunderBar

Lucy Carey
w3

The story behind the device blending software and hardware in one chunky package.

The WunderBar is designed to be a simple launchpad for
developing apps for the Internet of
Things
– without needing to learn about hardware. It‘s got a
clever chocolate-bar like design which allows sensors to be broken
off and positioned according to your own personal specifications,
paving the way for a myriad of potential IoT apps – from policing
your cat door to monitoring the temperature of your baby‘s
room. With a crowdfunding campaign to fund the device
recently closing to the tune of $111,472 – over $20,000 more
than the initial $91,000 goal, European startup Relayr are
riding high on a huge wave of interest. In this
interview, we speak to Jackson Bond, Co-founder and Head of Product
at Relayr and Chief Engineer Paul Hopton for the full
story on this aesthetically pleasing assemblage of software apps
and discrete BLE sensors.

JAX: What was the inspiration behind WunderBar
– did you have any involvement with other IoT projects prior to
this?

Bond: We had the idea since
spring 2013 to start a cloud platform as a service for the Internet
of Things. But we struggled to decide how to bring it to market: a
vertical or horizontal approach. Should we select an industry, like
home automation, health care, logistics, etc. and launch a solution
into a competitive market space? Or should we launch horizontally
and offer something for all developers? In September, we were
accepted into Europe’s leading Startup Accelerator, StartupBootcamp
(SBC), in the Amsterdam program. We had one goal at SBC, to figure
out our go-to-market strategy. There we were surrounded by hundreds
of mentors and experts in all aspects of product launching, design,
hardware, software, marketing, etc… and with the help of the
mentoring, and with rigorous customer discovery at app developer
meetups and hardware/maker meetups, we discovered the real problem,
that both hardware and software developers are lacking the simple
tools to connect to each other. This is where the Wunderbar was
born.

The hardware guys have Arduino and Raspberry Pi,
as “easy” starting points, but the software guys, the app
developers have nothing. So we thought about how to bring sensors
and wireless technology together in one easy kit. We first thought
about offering a wide selection of sensors, like a Forest Gump box
of chocolates. But when we asked the community about what types of
sensors they would like, it got too complicated. Everyone wanted
something different. So we took only the top six sensors and
connected them into a kit. We first called it KitKat, but Google
had just named their newest operating system KitKat, and we did not
want any problems. And so after some late night brainstorming with
the other teams in Amsterdam we came up with the
Wunderbar.

You’ve said that WunderBar is a sort of
hardware ‘hybrid’. Can you explain this in more
detail?

Bond: Hardware hybrid?
That is slightly misnamed. We approached hardware with a software
developer in mind. Its hardware, but we tried to make it fun and
approachable for people who have no hardware-knowledge. So they can
program the Internet of things within minutes out of the box. We
are focussing on app developer tool paradigms like
XCode.


What challenges did you face in putting the
whole package together?

Bond: We
from the founding team had limited hardware expertise ourselves. We
come from software and mobile development. But many tell us this
works to our advantage.

Can you give us a technical deep dive into the
software behind WunderBar?

Hopton: There is a lot
going on under the hood. We have probably nine different sets of
Firmware running on the wunderbar itself – all written in
Embedded-C. We use the Bluetooth 4.0 stack to communicate between
the Sensors and the Master Module. The Master Module communicates
with the sensor data and receives commands over MQTT (a light
weight M2M protocol over TCP).

On the platform side we are building a number of
APIs based on Scala and Akka, but with some services written in
Node.js. We are also evaluating various NoSQL data storage
solutions for handling large data-volumes – today’s favourite is
Riak. To provide A good user experience we will be building
reactive user dashboards in javascript.

We have Android Libraries and iOS Frameworks to
do some of the heavy lifting for app developers, as well as to
enforce some security constraints. We will also be open-sourcing
libraries for other languages such as Python and
Node.js.

What are the coolest/ most significant
real-life implementation examples you’ve come up with so
far?

Bond: We have not focussed on
developing single implementations. We don’t believe that we will be
the ones to create the IoT killer apps, so are trying to enable
developers by bringing the barriers-to-entry down for developers.
We do however want to show consumers that even simple problems
could be solved with an easily available sensor. Smart solutions
should not cost 4-5 figures for your home or health care or
otherwise. We think that once folks can buy a smart solution for a
a few dollars in their hardware store the age of the Internet of
Things will finally have arrived!

We have experimented with a couple of
ideas:

  •  Health Care:
    connecting different existing devices (connecting existing
    bed
    sensors, motion sensors, smart wristbands, etc)
    for better
    supervision for elderly at
    home.
  •  Security: simple
    subscription based home solutions with a wifi camera and a
    contact
    sensor, sharing images of intruders with your
    trusted network and
    authorities via mobile phone, in
    case you are not at home, nor in
    the
    vicinity
  •  Fun App: see
    our video: You have Snail Mail App, using the

    light/proximity sensor, to let you know when the mail
    arrives.

Do you think efforts on the part of Eclipse to
avoid fragmentation within the IoT will be successful? Do you think
more could be done?

Hopton: Ian Skerrret, Benjamin
Cabé and the rest of the Eclipse IoT group are leading a really
strong effort. We are using MQTT and the Paho library for
communications between the WunderBar and the platform, we also
expect to be contributing back. Interoperability is such an
important topic for IoT that people developing platforms should
utilise existing platforms where possible. A part of me remains
sceptical – on a practical level there are so many concepts of what
the IoT is, and how it should work – there can be no single
platform at the moment. I do believe that well engineered
components like MQTT will see wider adoption. and that is a good
thing.

Oracle is positioning Java as the platform
that will ground the internet of things – what is your opinion on
this?

Hopton: Java as a language is
showing it’s age, especially with the framework-itis that has
defined J2EE development over the last few years. Younger
programmers are just finding dynamic languages more attractive.
That is why we are seeing an increase in the number of platforms
running interpreters for Dynamic languages like Node.js or
Python.

Probably the core issue is one of image. Many of
the bigger trends in IoT are Open Source and I don’t see much good
will towards Oracle from the OSS crowd.

Finally, congratulations on such a successful
tech Kickstarter! What do you think is the ‘secret sauce’ behind
getting people to back your project?

Bond: Secret Sauce is good
social media and media preparation a couple months in
advance

(see my article on our crowdfunding experience)

and having a strong network effect at the beginning if
possible. We were helped by Paul Poutanen, Founder of mob4hire. He
wanted to offer his network of 60k mobile developers and testers a
first shot at getting the Wunderbar, so we chatted with him, and
really liked what he is doing for the mobile world, and offered to
give his community first access to the best prices for the first 4
hours of the campaign, before we publicized it. Paul sent out an
announcement to his “mobsters” with the special offer, 24 hours
before we launched alerting them to the deal. That really helped
give us early traction from around the world, I think we got about
25 backers from the mobsters. Without the “mobsters”, we would have
been so successful. Another luck was that we went with Dragon, a
much smaller, selective and niche site, which is watched closely by
the industry. 4 global distributors for electronics saw us there
and approached us. One of them, a big European media cpany, also
volunteered to send a mailing to their 150k electronics readers, in
exchange for a special price on the Wunderbar. This also brought
about 60 backers. Had we been on Kickstarter, we would not have
been noticed among all the other hardware projects.

We are still taking pre-orders
at:
 http://relayr.io/wunderbar.html

 

Author
Comments
comments powered by Disqus