feedback from Open Data France

[Natalia Martinez-Winter]

Hi,

I'd like to share with some interesting feedback I got recently when we hosted an event this week at Mozilla Paris for Open Data France.

It's a group of local authorities (regions, cities, counties...) their goal is to open all data that is generated by them. I felt Sensorweb was an obvious project to discuss to understand the challenges of open data.
They shared with me their experiences with environmental data collection (noise, allergens, pollutants, PM2.5) :

• devices: some tried to make devices very affordable in order to deploy in masses. But they realised that the generated data was unreliable, so the number of sensors wasn't making any difference. -> important to compare data captured by sensors (especially when cheap) with "reliable" (existing) data
• sensitivity of environmental data: depending on how the data is communicated, citizens may over react. In some cases, they've noticed psychosomatic effect.  -> need to manage expectations and avoid creating panic.
  
• relevance of processed data (communication issue): oversimplification of the results (scale from 1 to 4 for instance) doesn't help users in understanding the data -> important to provide meaningful reports and clear maps
  
• conflict with governmental policies: the typical scandal of Tchernobyl (the french government said that the cloud never crossed the border, all data was censored) -> good to be aware of that. makes it even more relevant to crowdsource data generation and have independent communication platform
  
• feedback from experiments with costly projects: local authorities partnered with big companies (Orange, Veolia) which provided the solutions/software to collect/process the data. In the end, the data belongs to them! -> now they learned the lesson and will be careful in their next contracts negotiations. Their request is for databases to be under DBL license. BTW what's the license of collected data in Sensorweb?
  
• a representative of a relatively small city in rural area expressed great interest for the Sensorweb project, he said that even if it's expected that the air is very pure in their area, citizens are likely to ask for reassurance of that. Having facts & measurements that prove the good air quality increases attractiveness of the location.
• a representative of the 2nd biggest city in France (Lyon) and one that has a reputation of being very polluted is running many initiatives : they're launching sensors (weather station, pollution, PM2.5) and a platform in 10 high schools. The platform is available for kids so they learn how to analyse and understand data (it's now in the school program).
  
• challenge: large areas (covered by costly sensors) versus hyperlocal data (from cheaper sensors): pros & cons for each. maybe the future will need a combination of the 2 (today it's totally separated)
  
•  question: value of mobile sensors ? examples: buses, bikes, pigeons
  

 I think Open Data France would be happy to collaborate and even partner with us so please let me know if that's of interest to you whether you want to get more insights, feedback from former experiments or want to select projects, locations (cities, rural) to betatest Sensorweb depending on your needs.
thanks!

--

Natalia

[Sandip Kamat]

This is useful, thanks Natalia. From platform perspective, these will be useful guidelines on data publishing and visualization. Potentially some of this input may go into what we want to ensure before we certify projects with our platform.

btw, I am hearing some interest from India as well as people hear about sensorweb. Air quality is a big issue there.

Sandip