News‎ > ‎

20111107 Protei general meeting notes

posted Nov 9, 2011, 6:25 AM by Cesar Harada   [ updated Nov 9, 2011, 7:32 AM ]
[In progress - editing]

We have no had a meeting for a very long time since summer in Rotterdam, so this was really to know what everybody's been up to and what have been individual progress. We'll hold the next general meeting next WEDNESDAY Nov 16th at 16:00 Oslo Time, same place (Skype). For different timezone, please check :
So, here are the individual updates. 


1. Etienne is working for a proposal with TU Delft and Southampton University with Olivier. 

2. Etienne is working with Paul + Seb (APC) on composite materials. <cesar> I got really good contact with Pierre Pomiers of NOTOX - it would save us a lot of time to go do a field trip to his workshop and learn about the materials and techniques they use. He's interested in a partnership I think </cesar>

3. Etienne mentions Sam Nemeth and Koert van Mensvoort (TU Eindhoven) doing some reseach with us through the Next Nature program:

4. Etienne has been working with Peter, Gabby and Zenon in Sydney regarding Artificial Intelligence research program. The first part of the research would be to have 
- an AI for one Protei
- an AI for a swarm of Protei

5. Etienne suggests the following order of work : 
    1. identifying the concrete missions for each application of Protei, starting with oil spill collection
    2. design the research questions that we need to look at to be able to carry out these simple missions. for instance: how much load can we pull? how much do we want to pull?

6. Etienne reminds us we have very good connexions with many famous universities that are eager to work with us if we are smart enough to articulate our research questions : Architecture institute, Stanford, TU Eindhoven, Ecole Centrale Nantes, Ecole des Mines, etc.
So what we need to give them is research packages : 
package = 1 research question +  how it is linked to the overall picture of the project

7. Etienne suggests the formation of subgroups for in depth discussion. 
groups : 
1. Group Admin, communication Startegy. _ subgroup lead : Cesar
2. Group Mechanics _  subgroup lead : Cesar
3. Group Electrics, Electronics _ Who would lead this group?
4. Group Software,  AI _ subgroup lead : Peter 

<Cesar> About 6. and research packages : to clarify  I would suggest to 
- focus on oil spill collection (it is by far the hardest mission I believe). 
- Describe Protei in A FEW critical questions
- Take these questions as "research packages" that we can give to Universities (prof / students) as a brief. This brief should contain some experimental setup suggestions (i.e. towing tank machinery, sensors, and range of testing to re-create natural conditions) </Cesar>

<Cesar> About 7. and subgroups : we need to have one person in charge of each of these groups else it won't work. We need to have meeting notes of each of these groups = that is the responsibility of the subgroup leader. In an ideal world, the general meetings don't need to be attended by all members but by sub-group leaders : so we save time and we clarify the "chain of command". The general meeting always being accessible to everyone of course - especially if major issue arise. </cesar>

Piem Wirtz

is going to visit the RDM campus in Rotterdam with the Dynamo People. 
Piem has many questions : do we plan on one fullscale? Several smaller etc?
Piem needs to know if we are going to all gather this summer in Rotterdam again, perhaps at RDM. 

Piem stresses the fact that it is very productive to work together over short times (like we did this summer) instead of having multiple places of production. 

<cesar> I would agree with that, and I think having multiple work places during the year leading to a general summit during the summer is a great way of doing it. So basically we all work where we are and we gather all during the summer for a great Protei camp! </Cesar>


Alvaro suggest we take as first research question (for Etienne) the bollard pull of Protei. 
On top of this Cesar added, this is the question asked by and Etienne added "for oil spill application we need to know what´s the resistance in water of a sorbent boom saturated with oil"

Guys this is my proposal we may pass to the universities, I believe Hydrodynamics teachers will like it.

Bollard Pull contest
1- We can test with a reduced model and with calculations estimate the final power
2- Will define the size of the tail and equipments (e.g. radioactivity measurement equipment)
3- Simple static test (numbers!!!)
4- Need just a small tank , a load cell and a ventilator, 
5- May test it with several different sails.
6- Can work with Universities, the best relative size  (we may use international sailing rules to define a correction factor) versus the pull power /wind speed, receive a prize for example.
At the end we need power not speed for Protei and business people need numbers.

For example at the end we could say with a 3 meters model, 10 m2 sail, 10 knots wind we can pull 200 kgs.


First time for Aurelie in Protei workgroup : 
Aurelie is 33, been living in Paris for 4 years after 4 years in San Fran, CA
Aurelie is running her own consulting company called Embody.
Aurelie met Cesar at TEDx Basque Country (south west of France)
Aurelie has Been involved in the TEDx organisation in France. Aurelie has some connexion for the Protei Project in France and the US and would like to help us finding sponsors and €€€€. Aurelie is excellent at finding money, organising event, communication, and having a strategic view
Aurelie has been presenting the project here and there and says that Protei interest VERY MUCH big companies in France e.g: EDF/ AREVA/ also GENERALI (Raceboat Sponsor) the AMERICA'sCUP and so AND what they need :
- 1. a 3 years Business Plan view
(including a map of the team involved and the parnerships)
- a clear ROI (Return On Investment) 

Aurelie suggests it would be rather difficult to have big corp to support an R&D projects, but to give them ROI one option could be: giving them media visibility. So that they don't have to take part of your strategy and influence the direction of Protei.
So Aurelie suggests : an annual Protei Event could be a solution.
It could also be an opportunity for you to make so tests and collect datas as well.

<cesar> Aurelie and I have been looking at San Fransisco as a very good place to make a Protei major event. San Fransisco has the silicon valley + the best sailing scene in the USA. Also other reason is that we may want to invite all the biggest players / best sailors to be there at the launch. 


Might have found a great new place for Protei in Berlin 
The place is wonderful but need to be built pretty much from scratch + the winter might be cold. 

Suggests that there is 2 different areas of study :
- the design of Protei itself
- the different applications of Protei as a new technology. 

Need to clear the group mission / philosophy


Has been working to make the AI group happening with Gabby, Etienne, Zenon, Alex (Southampton) and Jerome (Denmark) to get the AI and control research group happening.  We're currently on the "Terms of Reference" up for discussion.


  • Protei differentiates itself from other robotic sailing endeavours because it seeks to develop a system that combines multiple vehicles in a co-ordinated group.
  • Protei seeks to develop modular components of a multiple vessel system that are themselves competent and who can act in concert within a heterogenous distributed group for a common purpose.
  • The Protei project also involves the assessment of flexible sailing structures to investigate if they provide improved maneuverability and can act as efficient wind powered towing vessels for a variety of uses.
  • Protei has an initial focus on the collection of downwind oil sheens and consequently has an interest in the behaviour of towed sorbent booms under a range of wind and sea states.
  • Protei is an open source product and as such undertakes to provide information freely to collaborators and developers in order to improve on the initial concept within constraints imposed solely to facilitate this development and maintain compatibility across developments if required.

Intelligence and Control:
In pursuit of these aims the Artificial Intelligence and Control (AIC) research group will seek to develop algorithms and control features through:
1) Develop a competent rigid hulled sailing vessel on a small scale suitable for testing control algorithms
2) Using this platform test and refine algorithms that allow the vessel to perform simple goal oriented navigation tasks and successfully tack across wind using it's control surfaces
3) Develop models for distributed networks of autonomous units through academic programs testing these at smaller scales until reliable then scaling up to the real world.
4) Pursue development of collective intelligence systems allowing a high level of self organisation within the distributed network and group competency without the need for intensive external control.
5) Develop communication systems that provide for robust linkages between individuals and a remote control centre utilising existing infrastructure as much as possible.
6) Produce front end mission planning and control software allowing the easy configuration and control of a fleet of heterogenous vessels over a wide geographic area.
7) Conduct academic research in to the coupling of a flexible hull and it's control algorithm to produce a mathematical description of the vessel's physical state space.

Role of the Group:
The AIC group will pursue these goals by:
1) Identifying and defining research topics that conform to, and advance the aims.
2) Seek and engage appropriate researchers and collaborating bodies to conduct these topics 
3) Act as a co-ordinating and advisory body for AIC research
4) Develop recommendations for standards to ensure the compatibility of compliant hardware and software
5) Periodically conduct progress assessments and gap analyses 
6) Oversee and develop the AIC technology development plan in co-ordination with other aspects of the Protei technology development plan.
7) Provide advice and support for group members in their research activities.
8) Identify and pursue funding opportunities to support these activities and the continuation of the group

The University partners want a certain degree of autonomy and are on the other hand happy to share outcomes that address our own requirements, in some cases they are happy to directly address our questions through taking on student projects.  
The next step in some ways is to find a way of supporting them more directly.  Once we have more concrete questions we can start identifying research funding.  Once the group has a bit more momentum

Cesar brought up some references :
mowing the lawn (scan survey routine) :
cooperative work 
swarm battle


== Ireland ==
ireland was great. I saw some more stuff re .science gallery
protei _002 is doing great, there was a lot of interest

== USA and 1m Protei ==
1 - I'll be doing some robotic boat building workshops with Protei sort of as the topic of interest , at the NY hall of science, for 2 days 4 hrs / day in december, Perhaps Eyebeam!
2 - Also i'll be talking about Protei in December at a talk that happens monthly called LUCID
3 - I've been talking with some professors and am starting to build articulated servo- driven robots , modular self assembling. I'llstart to document that .
4- People have contacted me with interest to get it in some of the canals in brooklyn. If I can do a workshop / event with a 1 m boat and get it in the water with an oil collecting tail I think it could be cool?
5- Also a man in Alberta has some applications he seems serious about - calm waters, thin oil sheens, could be a perfect spot.
6- I have had interest from a marketer to do 1 m DIY ones also, like, for chrismas time - for kids to start to use: so we have here different products
7- Although 1 m ones can be used for testing of performance if they pull small tails in dirty brooklyn canals?
8- I' m interested in starting to develop one sensing object / a swarm of those - but if I could use a prototype(say from _002) to build a few small models and get them r/c sailing in NYC I'd like to do that too

== New prototype ==
it required too much torque for the motors I had (i was using some servos)  - it is r/c and has cable actuators but I just got new motors to replace  (it's hanging vertically and moves a little but does not articualate) as much as i'd like.
I got 5 continuous rotation servos so... I'm working more from like, what application do I want, and how can I use little torque...You'd mentioned Protei_007 - I could start to document that as a smaller model with slightly different electromechanics
<cesar> build from the servos to determine scale (not from a general design ideas) </cesar>


For this who don't know Magda : Phd on Intelligent sails & Biomimetics
Magda asking about deadlines and research plans.

Olivier Taillard 

Has been added to Cesar's Protei mail list.

Francois de la Taste

Shares an interesting business plan

Cesar Harada

== Order of priorities I'm Protei's applications==
Order of priority of Protei's applications
1. Cleaning oil spill
2. Sensing Radioactivity
3. Sensing / cleaning Plastic
4. Sensing fishes
5. Physical oceanography

== Production spots ==
Workshop production capacity in London
And I have equiped workshop + technician available and willing to help here in the UK :
with a huge Laser cutter + 3d printer etc at Goldsmiths Uni :)
We have production capacity in different places, excellent.

So we have London (Cesar), Rotterdam & TU Delft (Piem), New York (Gabby), Southampton Uni (Peter, Olivier), Chile, and maybe soon Berlin (Sebastian) and Oslo (Etienne). Need to add this to the map. 

== 2 different products ==
We have in fact 2 main applications for Protei : 
- sensing
- cleaning

Sensing can be done with a much smaller unit than cleaner, perhaps 1 meter for sensing and 3 meters for cleaning. So it is seems that we have 2 distinct products that we want to have at the end of the day. 

== Reports on many meetings ==
1. Rotterdam : RDM campus 
            Very nice place to build Protei. Perhaps next summer :) We'd rent a space over there. 2

2. Munich : BMWi in Munich. 
            They are waiting for our business plan and ROI. They have a great power, we must produce an ambitious project that will give them amazing visibility

3. Leiden : Heerema
            They are a bit sceptic about some engineering factors, especially Bollard pull. They are engineers and wont invest in a project that is not making sense in the engineering realm. So it is good to discipline our project, but it limitates our research method (yes, there are other methods but engineering ones :) to be merely engineering approach. i think of design thinking approach. I think of experimental physics approach, that can be more creative, iterative and more about risk-taking than optimisation. 

4. TEDxBasques country. The presentation went very well. see it here:

4.1. NOTOX 
            Interested in working with us. They have in house the materials we need, they build amazing surf board in high performance bio-resins. We should spend a few days in their lab in the south west of France and establish a good partnership with them. 

4.2. Directrice Rivages Pro Tech (Pantxika OTHEGUY Lyonnaise des eaux) Water quality monitoring. 
            The director of water quality monitoring in this part of France is interested in using Protei for water quality monitoring. Better than that she gaves me the general specs of the equipments they are currently using on buoys. Even better then that she told me she was excited to put the equipement they have on Protei to get data!!! So that's extremely exciting and I think we have a partner who's already working in that space -> that would accelerate our dev so much! In a lovely location. 

5. TEDxSeeds (Yokohama) . The presentation was translated in japanese went very well I think :

6. FuRo. Thanks to Nae Morita. 
I met with key FuRo members at Chiba institute of technology. They have this amazing new arduino-type kit that they just released and that is very powerful : This a kit they give to engineering student to get familiar with basic electronics. They gave me one kit (!!!!) , and the idea is that we would send them some basic Protei and they would be happy having their student built the brain and guts of a 1meter Protei !!! FuRo is amazing. So let's build one protei 1m and send a few to them !

7. Miraikan thanks to Nae Morita. 
Same, this is an amazing institution focused on education that would be interested to promote protei, perhaps in their pedagogical programs. 

8. Japanese universities. I met with several professors from Toodai (japanese MIT), Kyoto Institute of technology, University of Tohoku ect and they were all interested by our project in order to do radioactivity sensing - it is amazing how Japanese are responding to our project. They have the understanding of the technology, the capacity to develop it, the urgent need to do so. 

    I met with one of the director of Qualcomm Japan and he's willing to equip Protei for telecommunication instrument for radioactivity sensing. I don't know if you realize but this is huge. If we pull Protei 1m soon enough in the zone, we can bring extremely valuable data.  

Both these 2 amazing environmental organizations have helped me make this map of japan coast : looking for the best spot to launch Protei.

11. Fishermen is Soma
I met and discussed in length with some local fishermen in Soma. They want to know the levels of radioactivity. Currently they are not allowed to fish and their livelyhood in in danger. I think they would help us deploy some radioactivity sensor very near the power plant.

12. TEDxOrlando : presentation went very well. 
Connected with the nice Local hackerclub. They want Protei 1m for christmas :)

13. VIDA
We have won that competion, so we may have shows coming up and opportunities from there. Protei is very likely to be exhibited at ARCO in Madrid in February 2012.


Apart from all that networking for Protei Cesar has been working intensely on the business plan and an expanded version of tech dev plan with milestones, deadlines and deliverables. Soon more to come :)