A fully autonomous digital pen for notetaking, drawing and learning; points out writing errors and connects via WiFi in real time Read more
This project's funding goal was not reached on August 9, 2013.
About this project
>> DEUTSCHE ÜBERSETZUNG <<
>> РУССКИЙ ПЕРЕВОД <<
We are two fathers (and a wonderful team) with one common goal: bringing more "good vibrations" to business, families and schools around the world by creating a unique digital pen that's both useful and fun.
Lernstift builds a bridge between classic handwriting and the possibilities of digital technology, creating various new ways of note-taking, digital drawing and e-learning. We invite you to build this bridge with us.
Welcome to our journey ...
A whole new kind of digital pen
Current digital pens use optical sensors to pick up the writing movements and digitalize the words or sketches for computer use. Some of them require external devices. Others only work with specialty paper.
Lernstift is different. It requires none of those. It's the first truly intelligent pen because all the tech and software is built-in. It can be used with apps but it doesn't depend on them.
Here's a video of our current prototype in action
Real-time WiFi connection with
smartphone, tablet and PC
Thanks to its built-in WiFi module, Lernstift can connect to PCs, smart devices and even other Lernstift pens. This opens ...
A world of mobile and PC apps around the pen
Imagine: Parents can check on their kids' learning success with a learning statistics app. Two (or more) Lernstift users can work together in a single document using a Lernstift co-writing app. Kids can play learning games together or send their grandparents handwritten postcards via Internet. Everyone can share their notes and sketches via Facebook or Twitter with a Lernstift app for social media sharing, and much more ...
Apps for school use
We're already in contact with education officials who are interested in integrating Lernstift in their curriculum and planning to develop school-based software that works with beamers and tablets or as an alternative to expensive digital whiteboards. This would, for instance, allow teachers to see, in real time, what their students are writing, and to respond instantly. Individual results can be shown and discussed (anonymously, if desired) with the whole class to make the learning experience more engaging.
To assure the quality of our educational software, we will work in close cooperation with schools, universities, educators and educational scientists. (As a first step, we've been offered a "testing class" plus scientific advisory council by a school and university in Salzburg.)
Open API for developers
To allow boundless creativity, we will open our software platform and allow access to the pen's sensor data and handwriting recognition. That way, developers from around the world can create their own Lernstift-based apps.
The pen's two primary functions
1. Orthography Mode – to recognize spelling errors
– for writing starters (5-8 yrs) + everyone (7-99 years)
2. Calligraphy Mode– to point out flaws of form and legibility
– for writing starters 5-8 yrs (7-99 yrs in Asia)
Two languages to begin with ...
And a lot more to come.
Upon launch, Lernstift will be available in English and German. As soon as we can, we will roll out more languages: Russian, Spanish, French, Italian, etc.
Embedded Linux inside
The computer inside Lernstift is an embedded Linux system. The board contains motion sensor, processor, memory, Wi-Fi and vibration module. It is being designed to fit perfectly inside the pen.
We are using an individualized board, based on the Gumstix Overo. To realize this specific board, we work together with a German company.
The sensor in detail
Our in-pen computer comprises a special motion sensor that combines gyroscope with accelerometer. To optimize the motion recognition we also add a magnetometer.
According to our calculations, the sensor delivers a resolution of ~400 dpi at 200 Hz. The handwriting recognition engine (more infos below) requires only 100 dpi at 50 Hz. In other words: We've got plenty of sensor power!
There are 3 different methods to process the sensor data for the handwriting recognition engine. With the help of our math guru Michael, we are now able to get the fine graded sensor information we need. One solution is a math calculation that generates a 2D path (the usual shape of handwritten letters). Getting that clear and precise 2D path is tricky because of sensor noise and a few mathematical "problems".
However, for optimal digitalization of fine drawings and freehand writing we have to get even better, hence make the generated 2D paths virtually error-free.
The handwriting recognition in detail
The pen's HWR software is not being developed by us. (That would be far too complex for us as a startup.) Instead, we are working with the worldwide market leader in handwriting recognition. We license their engine and extensive language database of over 40 languages. This will allow us to roll out more and more languages step by step ...
The engine is tried and tested on millions of tablets and smartphones around the world, but the application in our pen has its own requirements. These are our learnings from the first intensive testing:
The generated sensor data exceeds our requirements in terms of resolution and speed
The selected embedded hardware (ARM processor, 128 MB RAM) performs at the perfect speed
Recognition speed for block handwriting at 20 ms per character!
Our strategy for further recognition improvement
We will develop an optimal parameterization of the engine's API and use optimal the recognition result-tree. We also utilize the linguistic intelligence of the HWR better than other digital pens and tablets do.
In its first version, Lernstift's orthography mode will offer a spelling check based on single words. Through software updates we will later improve its capabilities step by step. The long-term goal: sentence-based grammar check.
Learning the user's handwriting
But even the pen's user help to improve the recognotion results: We will create a learning mode for the writers' individual handwriting style. To calibrate the pen, the user simply writes a few words in block letters followed by the same words in cursive. On the basis of this "base data" Lernstift can learn the characteristics of the user's handwriting …
Community-sourced extended word catalogue
Furthermore, we will implement a community framework / app: Through this, users can add their own private and public words to the pen's language catalogues. We will also provide an app that can be used to create punctuation and orthography rules. That way, each Lernstift user becomes a creator.
To ensure ideal ergonomics for kids, we're working with ipdd GmbH & Co. KG, an experienced industrial design firm from Stuttgart. They are developing the pen's ergonimic design.
Exchangable writing tips
We were able to realize an important product design goal: implementing a system of exchangeable writing tips. In other words, Lernstift will be a three-in-one pen: pencil, fountain pen and ballpoint.
Fountain pen and ballpoint will be realized with the help of the German pen manufacturer Online Pen who already produce and sell these exchangeable tips. We are still working on a solution for the pencil variant.
Doing the swap
The video above illustrates how easy it will be to change the writing tips on Lernstift. And, if you listen carefully, you might hear a familiar sound ... ;-) (The recognition is far from perfect, but we're getting there!)
Color variants for boys and girls
We're all about equal rights for men and women! But when it comes to colors, boys and girls may have different tastes. That's why Lernstift will be available in (at least) two colors:
Fully updateable software
Once our pen is ready for you, its hardware will remain pretty much the same. The software, however, will be updated and improved on a regular basis. As a consequence, even the early models of Lernstift will always remain up-to-date.
THE BIG PICTURE
What makes Lernstift so unique
You might know the Livescribe pen by Anoto. Their technology is based on specialty paper (pre-printed with an encoded pattern) and a small camera inside the pen to determine its position and movement on that paper. With their pens, Anoto has built what we call a “product platform”; a broad range of possible applications for that technology. But their platform is 10 years old and has a number of important limitations. (Take away the paper, for example, and the pen becomes useless ...)
This is where our big picture begins: In addition to our pen (the "educational vibrator", as some people call it) we are creating a whole new product platform. As soon as the pen is fully developed and launched, we will roll out a number of other products for educational, business and home purposes; all based on the same technology. In other words:
Lernstift is a lot more than just a digital pen ... It's a world of possibilities waiting to be explored ...
We will also provide an open API, so developers can create Apps around our digital pens, using the sensor and handwrtiting recognintion data that will be provided in real-time via WiFi.
Also, we are social media lovers. So we will make it possible to connect the pen and the Apps to your favorite social media networks – Facebook, GooplePlus, Twitter, Yahoo, etc.
What we've "built" so far:
- Prototype housing and prototype computer were "married"
- Motion sensor perfectly connected
- HWR software runs fast (~20ms per character in block handwriting)
- Sensor recognized his first words: "Mama" and "Papa"
- PCB design started (oval shape)
- Inner construction of the housing is ongoing
- Kickstarter project ready to launch
- Investment negotiations with a major German game manufacturer
- Connections to retailers and distributors have already been made
So far, the evelopment is going perfectly according to schedule!
How our project has been financed until now
We are very fortunate to have received great support from a number of private investors (generated with a DIY crowdfunding through our own website). Their investment (and in some cases active participation) has brought us to where we are now. So here's a big fat thank you to them!
We'd also like to thank all those who wanted to become early investors but couldn't be incorporated.
Through our "DIY crowdfunding" we have been able to generate investments of nearly 200,000 USD from German, Austrian, Brazilian and British individuals and companies. And: we already have over 800 pre-orders!
What we want to realize with your support:
- internal construction of plastic shell (June 2013)
- ergonomic testing of pen's body with kids
- implementation of handwriting recognition with onboard software
- development of language exchange onboard software
- merging of body and prototype (August 2013)
- alpha testing, inhouse and with local schools (September 2013)
- production of pens for Kickstarter backers!!! :-) (October 2013)
(All dates are estimates/goals that might not be reached.)
How we'll make it real
We need about 600.000 EUR (all in) to reach the serial production stage. Besides the investments of private investors, we also have a few major corporate investors. This brings us the possibility to plan save till Q1 2014. But the absolute best case would be when we all together create the Lernstift with the power of you and the crowd!
We're glad to have strong partners on board to work on many important topics parallel.
We work with ipdd from Stuttgart to create a perfect ergonomic design for kids an adults. These guys have won many Red Dot Awards for their product designs. They have also created fantastic pen designs for Lamy (a renowned pen brand in Germany).
Global operating Consumer Goods Company (France)
With this Company we are working on the inner construction of the Lernstift pen. We are also establishing a strong cooperation with them in terms of production, distribution and the creation of Lernstift accessories. In addidtion to the pen's CAD construction this Company will also produce our first mini-series of plastics (silicone molding).
The handwriting recognition is naturally one of the most important parts. Here we decide to use the Software of the definitive market leader Visionobjects. So we can save time and be focused to build a perfect product instead own mad science on handwriting recognition. We already have tested the Solution and it works like a charm. In this days we use their API to write our Lernstift core application.
Lernstift Inhouse Team
All the software (besides the handwriting recognition) is created inhouse at our headquarters in Munich, Germany. Our geniuses are Peter, Thomas, Michael, Ravi and Adham. (More are to follow). The software development is the core of Lernstift's application framework. We divide it into these sectors:
– Sensor development
– Handwriting recognition development
– Core embedded app
– Developer API
– Mobile app development (Android, iOS and Windows Mobile)
Development schedule (from Dec 2012)
We work along a project plan from December 2012 and we are happy to have reach all our goals on time.
Phase 1 (Aug/Sep 2013)
– First 20 to 100 units
– to collect early adopter feedback and refine the product in short cycles
Phase 2 (end of 2013)
– 500 to X.000 units
– to serve Kickstarter backers (product Version 0.9)
Phase 3 (end of 2013 / Q1 2014)
– serial production
A passionate bunch from around the world
We are a like-minded team of pros from different areas of expertise, working together towards a common goal. Our headquarters are in Munich, Germany. But we also work in/from Hamburg, Salzburg (Austria), London and even the Philippines ...
Some of our team are also among the first investors, so you can imagine how deeply invested they are in their work ...
Falk Wolsky (Founder, CEO)
Entrepreneur, software developer and soon to be father of three. In terms of information technology the self-educated inventor is always „hands on“. He has worked for many different clients of various branches in the area of digital content processing (e.g. IBM, MontBlanc, TUI, LindeGas, MediaSaturn, Douglas and Microsoft). Not only does he have broad know-how and abilities with regard to digital content, e-commerce, mobile apps and social media, he also has a good basic knowledge of entrepreneurial matters. And – perhaps most importantly: He knows his abilities as well as his limits, hence is never too proud to delegate when necessary or advisable.
Daniel Kaesmacher (Co-Founder, CCO)
Advertising merchant, freelance conceptual copywriter, father of two. He learned his trade in the “1. Bundesliga” of the German advertising industry, working for national and international clients of various trades and sizes. Since March 2009 he has been freelancing for agencies and his own clients. While he enjoys working for established brands from around the world, his special passion has always been working for startups from around the corner. When Daniel learned of Falk’s Lernstift project he not only offered a pre-seed investment but also his expertise in the field of advertising and corporate communication. Soon, a congenial and productive teamwork was established which resulted in Falk’s offer to join the company as a co-founder.
Jessica White (Project Management, PR)
Jessica is responsible for all our Kickstarter-Project related communication. She will answer your questions and is your first contact point. Jessica is an art education facilitator, curator writer and founder of Thinc, an intercultural art education association. She studied at the University of Oxford, in Manchester and Salzburg where she now lives. She joined Lernstift to continue her passion for technological developments in education.
Alex Hanke (Quality Management + Customer Support)
Alex started his carrier over 10 years ago as a consultant for professional services in the new economy market. Since 2006 he has been working as a freelance consultant for quality assurance and project management for several companies. When Alex learned of Falk’s Lernstift project he offered his expertise in the fields of quality assurance, customer services and project management.
Peter Hüwe (Head of Embedded Development)
Peter is one of the 20 most active Linux kernel developers worldwide. He works at Infineon’s security hardware division. His knowledge and fast forward-thinking work style are great assets to Lernstift. In October 2013 Peter will start working for Lernstift fulltime.
Thomas Hüttenhofer (Embedded Linux Developer)
Thomas has a strong background in embedded systrems and microcontrollers. He also used to work for Infineon. Now, he develops the hardware drivers for Lernstift’s embedded Linux system.
Michael Fiegert (Head of Sensor Development + Math Genius)
Michael is an math genius and a crucial consultant for our software development process. He gives us broad and valuable experience in sensor programming. His expertise is based on 15 years at global operating company, working with sensor data and robotics. Michael is also among first Lernstift investors.
Ravi Raj (Embedded Linux Developer)
Ravi studied math and computer science in France and India. The math specialist and talented embedded developer joined our team straight out of university.
Adham Badr (Mobile App Developer)
Adham has long experience in building mobile apps on various platforms. He created a public elections app for Egypt with over 1 million users, for example.
Aunes Malugao (Assistance + PR)
Aunes holds a Bachelor’s degree in mass communication and has a lot of experience in the fields of social media marketing, telemarketing and virtual assistance. She assists the Lernstift team with various office management, research and PR activities.
A family product, inspired by a family moment
In the autumn of 2012 Falk’s wife had a flash of genius while she was helping her son with his homework. (He had been making a number of mistakes without noticing...)
"What he needs is a pen that points out errors right when he makes them", she said.
"Let's build it!", Falk replied – knowing as a developer how to make it happen. He founded the company, put together a team and soon found first investors. A lot has been achieved so far, but a lot more is ahead. That's why we're now saying:
"Let's build it together!"
With your pledge you can be part of a true innovation and worldwide novelty. Our goal here at Kickstarter is to realize this amazing product with the help of people like you, who believe in the idea.
Our goal and philosophy:
Risks and challenges
Our major challenge is overall usability. We need to make sure the pen vibrates at the perfect time to ensure it doesn't annoy or distract the user. We'll have a whole lot of intensive testing to do. We take time for this from august to november 2013 with about 200 Testers.
– energy management (ensuring long battery life)
– to make drawings possible (exact positioning in long time distances) is a hard task and will take time.
– adaption to regional / local language variations with community knowlegde
– smooth and controlled company growth
– international logistics and good warranty management
– optimal developer APIs
– as few hardware cycles as possible
No, in this first version the Lernstift pen will not have pressure sensors.
But we have it already in mind. How to integrate the sensors in combination with a exchangeable writing tips system is a miracle in the moment. But i think we find the way. In every case it is not yet the focus...
We like the idea of pressure sensors too for giving childrens a feedback how strong the right pressure on the paper will be...
1) A basic paper-to-device (note-taking) solution will be one of our first apps, yes. We have to decide the notes format on the Mobile or PC side. It is important to know that we just support text first. Later we will add drawings too.
2) We like to create plugins for sample with Word, Powerpoint, Photoshop and Illustrator to create the recognized text from the pen in the documents in realtime. So we start with a word-plugin first.
Yes. We will create apps for Android, iOS and winMobile. We also create later Apps and Plugins for standard Applications like Word, Page etc. on Windows and Mac. The second step will take some time so we just talk about here of 2014...
No, we will deliver the Pen with a english version. We will prepare english for US and UK as different versions (eg. word catalogues). Then we offer the download of other language catalogues over our website and the mobile apps. With the Lernstift standard mobile App you can "switch" then the language on the pen (internally by replacing the catalogues). The pen will support one language a time.
Generally we can use languages out of the following list... But we have to check and optimize the Lernstift for each language. This will take time and ressources. So we decide step by step which languages we support based on regional product launches... Naturally we will start with english, then german (bit of patriotism) then french, spanish, italian and russian... This will be hard enough for some time...
Here the list of possible languages:
Chinese Traditional (Hong Kong)
Chinese Traditional (Taiwan)
English (United Kingdom)
English (United States)
We like to realize with Lernstift a clean understandable product. So we simply start with the 2 functions. Regarding the ortography mode we start with vibration at spelling of single words. No grammar or punktuation at the Kickstarter charge. But! You know - everyone can update the Lernstift software on the pen with our smartphone app. So everyone can be still updated (with the same hardware). In 2014 we publish then a service for the community to create grammar and punctuation rules. We also like to add rules by ourselves but our resources per languages are not endless. So we think it should be useful to have an easy web / moble based process to add grammar rules for everyone.
We will have a API consists of some important features. All the communication from and to the Lernstift pen will done via JSON datasets.
We will provide all calculated sensor data and the recognition results (letters, words). Developers can access this data in realtime over the WiFi. For this we will have a small server, maybe nodeJS on the pen. (not finally decided)
We also will provide a another interface for setting all the configuration values like languages, vibrating timing and similar pen functions etc.
Instead the Livescribe Apps we see it as more interesting to have the Lernstift pen data within mobile apps. We will prepare some base mobile apps as playground including pen communication, sensor data samples and social media functions.
Is it a stand alone product or does it have to be connected to a computer / smartphone while writing?
1) While writing, the pen will be fully autonomous. No other devices needed. It's meant to be "just a pen" to make the use for kids as easy and undisturbing as possible.
But you can send the recognized letters and words, later drawings too via WiFi to a Smartphone or PC and do something great with it (note-taking too)
2) "Will there be any internal storage so the pen can work as a stand alone note taking tool?"
Hmm. We dont decide this point actually. There is much space on the lernstift hardware (2GB) so note-taking is not the problem. But makes this sense if you can have this data in realtime on your smartphone? We think we will have a survey to this topic. We can provide a solution if the demand is there...
We will deliver at least 2 writing tips with every Lernstift (fountain-pen and rollerball).
On the classic pencil sytem we still work because of its more fine mechanic components. Takes time ;-)
- (30 days)