Go Amiigo Activity Tracker Video review

Finding a waterproof activity tracker was an early requirement for the opensportproject.  And after some searching around the Amiigo.com  activity tracker was found on the Indiegogo crowd funding website.  The promise of an open source policy and database learning activity tracker all added up to make it a tracker to back and support.

A couple of weeks ago I received the tracker and took it to the swimming pool for a workout.  The video review explains more, but first impressions were positive.  It picked out the swim strokes and the timing were reasonable too.  It was comfortable enough to wear with a tight fixing, one notch tighter than I had when not in the pool.  Its not perfect, a bug uploading and analyzing the data has occurred on an ipad mini sync up. It is waterproof.  As for the learning, it needs to be taught kick and drills and more information is need to understand the form statistics etc.  Overall, well done on getting the innovative tracker to market Amiigo.

 

Posted in Uncategorized | Tagged , , , , , , , , , | Leave a comment

World Record Analysis 100m Freestyle Past and Future

The history of 100m Freestyle World Record Swimming goes back to 1905 with Zoltan Halmay recording a time of 1:05.8.  The woman records started in 1908 with Martha Gerstung recording a time of 1:35.0.

It is now 2014 and the chart below plots the record offical world records to-date for the 100m Freestyle for both men and women in a 50m long course swimming pool. The simple chart tells a story.

100mfreecompare

The most obvious trend is a downward direction in times recorded.  Bigger reductions in time further in the past and plateauing out of time since the 1980’s with a small dip around 2008 with the new swim suit technology. It is interesting that the reduction of male and female times follow a similar pattern albeit the female times are slower. But what about looking to the future?

Future Analysis
The chart on the right hand side projects/predict future world record times based on statistics. In this case by applying the LMS (least means squares).  This standard or the most basic of statistical approaches shows the trend of plateauing time reduction continues, but in a slight downward direction and that the difference between male and female world records has all but disappeared over the next 10 year.  I expect if the chart was extend we would reach that day where women over take male times.  Can statistic predict the real future?  Time can only tell.  Given advances in swim suit technology and currently females are allowed to cover more of their body then over the longterm this would aid faster time improvements as an potential explanation.  Might be women discover better ways to train etc.  Other statistical models have give similar outcomes for distance runners.  Some statisticians believe times will never reach zero thus LMS approach is misleading.  However, those approaches need to set the minimum absolute time a swimmer could not go below and that becomes a subjective choice.  All we know is that as more data is recorded, world records, we can review how well the statistical models are.

Posted in selfengine | Tagged , , , , , , , , | Leave a comment

Sport on the Blockchain

I recently wrote a personal blog post entitled, Putting QS on the Blockchain. QS meaning Quantified Self.  At the opensportproject.org sporting data is a subset of an overall quantified self.  The most well known blockchain is Bitcoin. While there is so much more to bitcoin than a currency, a cryptocurrency, a currency backed by the mathematics of cryptography. This ‘currency’ status feels like the core motivation of the community to-date that has been motivated to displace the existing centralised financial system.  We have exchanges, just like foreign exchange markets or event company share issues etc.  The fundamental difference between the old and the new financial system is one of centralisation v decentralisation.

Standing back and looking at the sporting world a large, it too currently operates around centralised Institutions.  Soccer – FIFA, Forumla 1 -FIA and in swimming – FINA.  The ultimate ruling body for ‘official’ rules and recording keeping etc.  But just like in Banking, there is no need for a centralised Trusted body to guarantee trust within the sporting community.  In fact, as recent and on going scandals, political in fight shows, these governing bodies are far from trustworthy.  In sport we do aim for a level playing field and a decentralised network of those participating in sport would provide a far more reliable, stable and democratic environment for sport to flourish, grass roots to world record holders.

Record keeping is a good example to focus on. What is a swimming world record?  Right now a world record is the lowest time recorded over a set distance for a set stroke, recorded on timing equipment certified to be accurate to a given specification.  The human identity is registered by joining the Governing body and the health, ie medical status clean on certain biological substances, simply put clear of banned substances. All this under the centralised governing body.  A couple of sentences written but with many knowledge definitions to be further explained:

Identity and health status:  A level playing field starts with the individual participating in sport.  Historically, competitive sporting bodies have firstly differentiated on sex but even this is becoming not so clear cut.  With a centralised rule book you adhere to the rules. Those rules documented in sporting Law and policed by testing etc.  How would a decentralised identity and health status be established?  The individual athlete would choose to put their sporting life “on the record”. The degree and transparency of the disclosure will be decided peer to peer.  Will that not lead to an un-level playing field?  Maybe, but it that not the case today?  A sporting event run on a self authored personal data would be a group of individuals agreeing to come together to race based on accepting each others status.  E.g. they could all use steroids or not, or in the future have gene therapy or not.  Sport is more than just the participants, it is the crowd or viewer too.  Again those individuals will have the same information as the athletes and will vote with the attention on what to watch, track or view.  The mechanism for all these individual peers to work all this out is the blockchain. It will not be time consuming as their own individual ‘smart contracts’ will be programmed to filter on their own personal preferences for participation or allocating their attention.

Recording the race, whether is be the time, distance or score.  Lets take a time – distance e.g. 100m freestyle long course ie 50m swimming pool.  How deep is the pool 2m or 3m, who did the measurements to certify this and what timing sensors were used?  Again, the centralised body rule book will set the standards and those will be guaranteed by the technology provider.  Again these sensors need policing and independently tested.  So, if a swimmer records a time lower than ever recorded in history before, they become the world record holder based on the accuracy of the sensor standards.  But does that mean they are the fastest in the world?  The increments in improvement are getting smaller and smaller, .01 even those times are recorded to 0.001 of a second, it just the accuracy of the sensor are not statistically reliable at those degrees of accuracy.  But the current touch/press based sensor can be replaced, e.g. by light ie  a photo finish with a high speed camera.  So, the world record in accordance or in the context of the sensor recording and our trust in them. These sensor will change with time.  While their is technical measurement uncertainty there is the much bigger question of who did actually swim?  Seems so straight forward to answer, the event organizer invited the swimmer to complete, the time keeper verified their identity and those watching or the recording of video provides evidence the race in fact did take place.  If such standards were achieved at any swimming race then we could agree the recorded time was indeed a world record regardless of the status of event?  Decentralised competition becomes possible as the sensors give more accuracy of the environment being played on is a level playing field.  Just because all swimmers are in the same swimming pool does not equate to a level playing field.  Competitive sport can be decentralised.

While competition and sport go hand in hand.  The meaning of competition will evolve too.  Again this goes back to the centralised rule book that states, freestyle technique is legal if it adheres to a b c of the stroke.  A judge, a human checks this is the case.  While us humans are amazing, our ability to see all swimmer at all time to ensure the entire swim is legal is beyond our attentions abilities. It is not for a sensor, video or smart water.  Human judges will increasingly give way to machines ie sensors plus software programmes.  Again those wishing to compete will choose the race judging rules and even the rules that the ‘winner’ will be declared, fastest time  or how a bout fastest time with pulse rate never going over 160 beats per minute, I am sure far better race rules can be dream-pt up.

A decentralised  blockchain swimming example, possible today:

Twenty swimmers attend their local swimming pool to train. A coach has prepared a training set (tomorrow a Machine Learning – smart contract will assist the coach or prepare it autonomously). The swimmers own various wearables to track activity and health vitals with an external timing system to keep all the wearables in sync and video analysis in real time from various cameras.  All the swimmers have their own tablet pool side providing analysis in real time.  This analysis based on their own data, peered connected data and the wider swimming sporting universe of data.  All the application run on e.g. ethereum.org platform on each tablet and with a maidsafe.net storage service distributes backup across the twenty swimmers tablets.  Some swimmers have their genomes analysed and a AI system is processing correlations, this take more than CPU power for one tablet so they use maidsafe.net compute network locally to share tablets processing power.  Once on the WWW, all data is synced and further analysed and preparation of future training session and predicted times for the next swim meet are visualized.

Posted in Uncategorized | Tagged , , , , , , | Leave a comment

Swimming wearables keep coming – kick

The latest wearable that is wearable has closed a funding round on Kickstart.com, Flyfit.  And ankle wearable thus allowing a swimmers kick to be isolated and data recorded.   With  GoAmiigo.com   shipping its waterproof wrist wearable we now have both upper body and kick covered.  When will a core wearable come along?

Posted in Uncategorized | Tagged , , , , | Leave a comment

Ecosystem overview

The goal of the OpenSportProject provide a range of connecting platform services, hardware to software to data that will allow the individual to quantify self their sporting world.

An overview of the current Ecosystem being put together.

And a summary of the technologies chosen thus far

Posted in Uncategorized | Tagged , , , , , , | Leave a comment

Sports Self Engine Flow Diagram

A bit of a geeky post today, a high level SelfEngine Data Workflow diagram to summaries the workings/logic going on in the software code of the Sport Self Engine.

Source code via the developer page.

Posted in selfengine | Tagged , , , , , , , | Leave a comment

Self timing and Identity Automation – trial

The new start and stop module was connected to the SportsPi and a bluetooth identity tag was place under a swimming cap.  From this setup the swimmer was able to start their own timing clock and see that times captured real time poolside.

This was just for one swimmer. Next stage multiple swimmers and capturing of split interval times autonomously of any human involvement.

 

Posted in Uncategorized | Leave a comment

Augmented Intelligence Pyramid

Augmented Sporting Intelligence:

Starts from the base characteristics of Robust and Reliable.  Works in all sports environments regardless of whether they be an indoor or outdoor environment and does so every minute of every day.

Data can then be Aggregated and from multiple sensors and sources to improve the Accuracy of capturing the sporting activity.

Quality granular sporting data can then produce Statistics that can be modeled into simulation of future sporting activities.

The ultimate goal to augment the individuals sporting intelligence, training to race day or  day to day lifestyle.

augmentedintelligencepyramid

Posted in Uncategorized | Tagged , , , , , , , | Leave a comment

Swimmer ID automation

One of the big goals of the project is to achieve complete i.e. 100% automation of the capturing of sporting data.  The initial sporting focus is data from a swimmer in a swimming pool.  But how do we know the IDENTITY of the swimmer?  Well, a coach or a time keeper can use their sight and knowledge to match them up.  What about removing the human from the equation?  The Identity Problem needs solving.

Computers and sensors are not ‘smart’ straight out of the box but they can appear to come to life via software.  The project has been able to do this.  Each swimmer wears a Bluetooth ID tag that emits a unique ID.  When a timing event occurs, e.g. start, tumble turn at any end of the pool or a finish.  The swimmer closest to the location of the timing event is allocated a time.

This setup was tested out at a real swimming pool this week.  There are still some distance communication issues to be addressed but overall the concept is working.  The data collected feeds into the swimming selfengine in real time to be displayed to the swimmers.

Posted in selfengine, wireless stopwatch | Tagged , , , | Leave a comment

Android Code – swimming sensor

The code for turning your mobile phone into a swimming sensor for Android has been published by Joseph.  GITHUB mobile swim sensing

Posted in Uncategorized | Tagged , , , , , | Leave a comment