Pachube - a detailed history of the early years

By Usman Haque

What follows is a history of Pachube.com, one of the earliest and largest open IoT data platforms, from my perspective as its founder, Usman Haque.

This document is an account of Pachube as it existed and operated in various forms between 2004 and 2011, from its conceptual beginnings in my London-based interactive art, architecture and design practice through to its acquisition by LogMeIn in July of that year.

It is not a history of what came after (subsequent rebranding and eventual sale to Google) though that is covered briefly at the end.

The document describes how Pachube came to exist, what I was trying to do, who used it and what they built; how it intersected with the open data and maker movements of the time, its contribution to the emerging internet of things industry, and the impact that it had in academic literature; in the community of practice around Arduino and physical computing, in citizen science initiatives like the Air Quality Egg and the improvised response to environmental disasters like the Fukushima radiation crisis.

It's a personal recollection, but I've tried my best to be objective and complete. Please let me know if you notice any errors or omissions, or leave comments at the bottom of this page.

---

1. Introduction

Pachube (pronounced "patch-bay" -- yes, this was always contentious) was one of the world's earliest open IoT data platforms.

It was built as a web platform for collecting, storing, and sharing real-time data from and between physical sensors, IoT devices and connected environments. It ran as an independent platform service at Pachube.com, from 2007 to 2011, when it was acquired by LogMeIn, at which point it was handling several million IoT datapoints per day from thousands of individuals, organisations and companies around the world.

The name came from the idea of an audio patch bay -- the panel in a recording studio or telephone switchboard that routes signals between devices. The analogy suggested a system where any data-generating device, structure or environment could be connected to any other, without either party needing to know in advance what they were connecting to. I was spending a lot of time in Japan at that time, so I had in mind a phonetic transliteration of the phrase (possibly also influenced by Sam Mulube's last name, since he was one of my first close collaborators!)

Pachube operated during a period when the phrase "Internet of Things" was moving from academia to something that technology designers, hobbyists, researchers, artists and then eventually companies were actively trying to build. There was no particular established industry at its public launch in early 2008 and few standard protocols for device connectivity or architectures for IoT data sharing.

Pachube was one of a small number of platforms that were created to provide that infrastructure, with a particular focus on open data, creative technology and community-initiated innovation. The patterns we tried to establish -- open RESTful APIs for sensor data, community-built app ecosystems, data organised into feeds and datastreams, human-readable metadata attached to every reading, machine tag triplets for interoperability, etc. became to a certain extent the template on which much of the subsequent public IoT industry was built. When AWS IoT, Azure IoT Hub, and Google Cloud IoT Core launched in the mid-2010s, they built on a set of conventions that platforms like Pachube had done much to establish.

---

2. The Origins of Pachube

Background

I originally trained as an architect and specialised in designing interactive environments and structures that respond to the people within them, where temperature, movement, sound, electromagnetic fields and other physical parameters could be sensed, processed, transformed and fed back to the environment itself in ways that the participants themselves could manipulate the transformation. I had been deeply influenced in particular by the cybernetics theories of Gordon Pask and his Conversation Theory, a framework that describes how participants (including machines) can create shared understandings without a pre-determined language.

My focus in the early 2000s was exploring architecture as an "operating system", incorporating elements of "hardware" (floors, walls, roofs) and "software" (sounds, smells, electromagnetic space, social programme), and in particular what "open source architecture" would mean in this context (both in terms of the design and the interactive experience of spaces).

Connected Environments

The first kernel of an idea that would later become Pachube was an interactive architecture proposition I developed in 1996, as a postgraduate student at the Bartlett School of Architecture, titled Conversations with a Kettle (and other friendly household objects), in which a domestic environment was populated with gossiping appliances using a language that humans could not comprehend.

This project was later broadened to conceptualise "connected environments", envisaging a system in which buildings, spaces and structures could share data in real-time, learning from and adapting to each other - which grew out of a problem that kept recurring in my design studio: each new project involving networked environments or monitoring sensors and actuators required building its own data infrastructure from scratch and, as other artists and designers started creating interactive spaces and publishing their data, there was no generalised way for one connected environment to share its data with others.

I was interested in how buildings could communicate with each other, strike up "conversations" and share information about their real-time occupancy; and how these connected environments would enable a globally-interwoven series of physical/virtual spaces. This wasn't purely theoretical: much of my design practice was founded on the idea of actually building things in order to test ideas.

Pre-Pachube Prototypes & Experiments

The idea that eventually became Pachube crystallised in part through a conversation at a conference in Riga, Latvia in 2004, RAM5: Open Source Media Architecture, where I briefly mentioned a working prototype platform providing "RSS feeds for environments" in my talk, after which Ben Russell (author of Headmap Manifesto, one of the earliest texts about location-aware computing, prescient about the implications of networked physical space) pressed me to explain more about the prototype and how it might work as a real service. I didn't have a detailed answer at the time but working with his question in mind over the following years contributed directly to the design of Pachube.

I continued developing what I was calling Environment XML (also known as EXML) throughout 2004 and it launched publicly on 12 February 2005, built on a LAMP server architecture, hosted as Environment XML within my design studio's website and open to other artists and designers to publish and 'connect' their interactive projects and environments (including objects and spaces in Second Life, a multi-user virtual world).

One of the earliest feeds was from iDAT at Plymouth University. The EXML platform included features that would several years later be a core part of Pachube: feeds, datastreams, graphs, tags, live/frozen stats, and multiple users.

Some of this period is referenced in an interview several years later. Environment XML saw usage by others grow over the next few months, getting up to about 70 artists, designers and hobbyists by mid-2006, so I sought other funding sources as the server costs continued to rise and I hoped to expand capacity.

First attempts at raising funding from art & design sources

In June 2006 I submitted a proposal to Eyebeam Art + Technology Center for a project titled "EnvironmentXML" which laid out the following problem statement for networked objects and environments:

"There is almost always a linear causal relationship between the input data (e.g. air temperature) and the output data (e.g. changing from blue to red) which means that the interface cannot be evaluated except at a purely aesthetic level or on a 'novelty'/'nostalgia' scale... It seems to me that rather than building another interface that responds to an environment, it might be more interesting first to construct a novel environment and open it up so that someone else might build an interface that responds to it."

The proposal outlined a system for streaming sensor readings from any physical space in XML format, tagged and geolocated, so that other spaces, devices, or websites might find and respond to them in real-time, geared in particular to architects, designers and artists working on interactive environments.

The document references Gordon Pask's Conversation Theory and second-order cybernetics as conceptual foundations, and cited as inspiration Stafford Beer's Cybersyn project (Chile, 1970-73) -- a networked real-time monitoring system for industrial production; as well as Julian Bleecker's concept of "Blogjects" (networked physical objects); Adam Somlai-Fischer's "Budapest Heat" project (which streamed a city's real-time temperature to a remote sculpture); and psychogeographical markup language experiments by Wilfried Hou Je Bek.

The proposal also described early conversations with Massimo Banzi about the Arduino ethernet shield then under development, and with Tom Igoe of ITP New York, co-founder of Arduino, who had recently connected an Arduino board to the internet using a Lantronix serial-to-ethernet converter. Another key reference was the work of Anne Galloway, who was researching the social and cultural dimensions of technology and design.

The proposal, along with several others submitted to other art, design and technology grant funders, was unsuccessful... but throughout 2006 the EXML platform's user base continued to grow; and so, as the server started to strain under load, I had to disable new registrations.

Meanwhile, moving between my own design practice and building the platform for others to use, several collaborations with Adam Somlai-Fischer had a direct influence on some of the decisions I made (and perhaps also on him launching his own platform which came to be known as Prezi). In 2005, I had collaborated with him to create an instruction manual "Low tech sensors and actuators for artists and architects" about connecting up devices to make interactive interfaces; we created the Reconfigurable House in 2007 using similar techniques, but the ambition was always to create a Scattered House as we did in 2008 -- environments connected and responding to each other in real-time over the internet.

3. Launching Pachube.com

Building the platform

In early 2007, I met Neill Bogie, who was to become one of the very first collaborators on building what was to become Pachube, at an Arduino workshop hosted by Alexandra Deschamps-Sonsino (then CEO of Tinker London). Neill introduced me to Sam Mulube, who started building the new platform 'Connected Environments', funded by my design studio, in October 2007, served at 'cnctd.net' as a Ruby on Rails/PostgreSQL stack.

In December 2007 'Connected Environments' was renamed Pachube, the domain was registered on the 28th, and launched publicly at Pachube.com on 19th February 2008, invitation-only, with a deliberately low public profile in order to prioritise stability before scale. The platform was designed "to facilitate many-to-many connections: just like a physical "patch bay" (or telephone switchboard) Pachube enables any participating project to "plug-in" to any other participating project in real time" so that, for example, buildings, interactive installations or blogs can "talk" and "respond" to each other.

The site, with tag line "pachube :: connecting environments, patching the planet", included tutorials for getting Arduino to connect to Pachube. There was no mention of the "internet of things", which came a few weeks later; instead at launch it said that the key aim of the platform was "to facilitate interaction between remote environments, both physical and virtual."

Pachube in beta

Over the next few months several hundred people applied to join the beta, the site design was tweaked (and also added the reference to 'internet of things') and by November 2008 there were around 200 active users, including researchers at Sun Microsystems, Telecom Italia, IBM, HP, and Alcatel-Lucent, and users at academic institutions across Europe, the US, South Africa, and Japan.

In early 2009 the team moved into a new office on Scrutton Street, London, in an area that would soon become known as Silicon Roundabout. We were initially in a building known unaffectionately as #shitoffice, but later we moved to 3 Scrutton St (where we stayed through the acquisition; and where my design studio remained until around 2022.)

I had a detailed discussion about Pachube and the internet of things with Tish Shute in January 2009 -- "Pachube, Patching the Planet: Interview with Usman Haque" -- which outlined many of my aims and ambitions for the platform, including how it related to my design practice (several of my interactive art, architecture and design projects were built on Pachube, with the help of Dot Samsen, who was by then working full time in my design studio).

We held a Pachube workshop at PixelAche in April 2009, and a short while later we were mentioned on the World Bank blog. That year, Bruce Sterling became a key advisor, Paul Pangaro helped craft the Pachube story and Rick Bullotta got so tired of giving us helpful suggestions in the community forums that we never implemented(!!) that he went off to found his own damn platform!

Early examples of things connected to Pachube included:

• The state of London's Tower Bridge (up or down), published by developer Nigel Crawley -- one of the more frequently cited early examples of a "thing" on the web
• Geiger counter radiation measurements from a Japanese laboratory (Kyoto)
• River flow and height sensors (Colorado River, Texas)
• A ship's location in the Pacific Ocean (Research Vessel Melville)
• Solar wind density from satellite telemetry
• A bedroom whose lights, fans, and heaters could be switched remotely by website visitors
• Second Life objects linked to gallery spaces
• Home electricity consumption histories by Matt Biddulph
• Building Management System data from the London Arup HQ published by Duncan Wilson

In the summer of 2009, a group of private investors from outside the UK, introduced by my brother, completed a pre-seed round of funding to help us support and scale Pachube as open IoT infrastructure, though still requiring invitation to use. I came to this entirely unfamiliar with how investors operated, and had no real understanding of the norms or how the dynamics would affect us. Over the next two years I had to remedy this naivety.

In December 2009, Ben Pirt, co-founder of Onzo, joined as Pachube's CTO, and guided by his scalability expertise we really started to pick up the pace.

Pachube open signup

On 21st January 2010 the platform was opened for anyone to signup (with dozens signing up that day alone). Several thousand technologists, designers, artists, developers and even companies joined the platform over the next few months, connecting up hundreds of thousands of datastreams from a broad range of connected devices and environments around the world. The rate of signups increased weekly until hundreds were signing up every week in 2011.

During the early growth phase many Pachube users were using the platform to store and share energy consumption data (including monitors as well as solar panel installations); later air quality monitoring was a key growth factor and also radiation monitoring (see below for more on these).

Much of Pachube's early momentum came from a small but enthusiastic community of people monitoring their own homes, in particular organised around HomeCamp, the London home-hacking and green-technology unconference, where Chris Burman and I presented Pachube's "open data for energy monitoring" at the very first event.

Pachube became a HomeCamp sponsor alongside Current Cost, Onzo and AMEE, and it was through that crowd that I first noticed people pushing their Current Cost electricity feeds into Pachube. To make that easier for Pachube users we built a Current Cost app and a CarbonFootprint app that used Gavin Starks's AMEE international emissions data to turn Pachube's electricity feeds, in watts, into a live estimate of their carbon footprint over the last fifteen minutes, hour and day.

By this point Pachube.com had a range of additional domains including international (pachube.cn, pachube.jp, pachu.be) and the international userbase started growing quickly as well.

A user survey conducted by Pachube around that time found:

• 84% used Pachube primarily for their own projects; 8% for company work; 8% for client work
• Most common uses: environment monitoring (62%), energy management (45%), tinkering and hacking (42%), home automation and building management (38%)
• By background: hobbyists/enthusiasts (34%), researchers/academics (16%), software developers (15%), product engineers/designers (13%), artists/designers (6%), company executives (5%)
• Most-wanted features: high-quality graphing/visualisations (62%), full history access (60%), data analytics and statistics (44%), open-source Pachube (44%)

Early team

The team was small. Our office manager was Emilfe Salinas, who worked across both my design studio and the platform. Ben Pirt had joined as CTO and from the start encouraged us to expand the dev team and hire a rock solid sysadmin, so it was a great relief when Levent Ali, Paul Bellamy and, soon after, Simon Forman joined us in 2010 (who also introduced me to climbing!)

Apart from Emilfe, Chris Burman was the only other person who worked with me at the same time both on Pachube and in the design studio. With this combined focus he had a particular insight on how to explore interesting uses for the platform from a design perspective (e.g. the first Pachube iPhone app, augmented reality prototypes and a plug-in for real-time data in Pachube to interact with SketchUp models.)

With my design studio work continuing to grow, it meant that I had, in effect, two full-time jobs (a working model I was to continue for another 10 years or so) and it's perhaps no surprise that my marriage was a casualty.

Platform, community, team and an industry all grow

The site got a shiny new logo (polished up by David Cuesta) and had a refresh towards the end of 2010, with the help of Joao Wilbert, who also helped put together an installation for London's Design Museum where the platform had been nominated for Design of the Year (interactive).

By 2011, users' connected devices and environments were sending more than seven million datapoints per day -- around 80 per second. For context, that was within sight of Twitter's messaging metrics at the time.

The Pachube team had grown as well, with Ed Borden joining in early 2011 to handle business/community development.

---

4. The Platform

Core Concept

The platform at launch was relatively simple. Anyone with a sensor or device connected to the internet could register a feed and push readings from one or more datastreams into it via a standard web API, using http PUT requests.

A key decision I made was that all metadata, apart from title, was optional and also user-definable (albeit with some helpers, like geolocation, description, exposure, disposition, etc.) This later generated some controversy (lots of conversations about consistency, interoperability and reliability) but I still believe it was key to enabling users of the platform to help define how it would be used, and what it would be used for.

Alternatively, for those devices that weren't setup to push data but were accessible over the open internet, Pachube could make regular GET requests (every 30mins) to its stored URL, retrieving latest datastream values. These readings were stored, graphed, and made available in real time to anyone else with access to the feed.

Initially all feeds were public, with Pachube's emphasis on building an open internet of things. However later, following considerable user demand, private feeds were enabled as well, which we realised might form part of a revenue model. Public data could be read by other devices, other people's code, or third-party applications built on top of the Pachube API; private feeds would be locked either to the uploading user, or the organisation that had provisioned the device's upload (in which case owners would be able to opt-in to making the data public).

MQTT and websockets were added later, also in response to user demand -- see more about this in an interview with MQTT's inventor Andy Stanford-Clark, another early Pachube user.

The system was designed around the assumption of continuous, real-time data flows, not static uploads. Feeds that pushed to Pachube (or that Pachube pulled from) were referred to as 'inputs', while URLs, triggers and external actions were known as 'outputs', expanding on the original concept of a data-distributing 'patch bay'.

Feeds and Datastreams

Data in Pachube was organised into two levels. A feed corresponded to a single source -- a building, a device, an installation, an environment. Each feed could have a geographic location, a description, tags, and privacy settings, marked either static or mobile. Within a feed, individual datastreams carried actual time-series data readings: one datastream for temperature, another for electricity consumption, another for CO2 levels, etc.

Each datastream had its own history, units, and tags (and later triplet tags aka machine tags, inspired in part by Aaron Straup Cope's work on this at Flickr), and could be independently graphed, subscribed to, or, later, used to trigger actions. Datastreams initially had numeric IDs and later alphanumeric were allowed as well. Premium features introduced the ability to batch upload historical data.

EEML

Pachube's original native data format was EEML (Extended Environments Markup Language), a custom XML format that grew directly from the earlier EnvironmentXML platform (and shaped, in large part, by Chris Leung, then a PhD student working in my design practice).

EEML was designed to describe the state of environments as they change over time, and to work equally for physical locations (a building, a river, a forest) and virtual ones (a webpage, a Second Life space); immobile buildings as well as walking cities and mobile devices. EEML included (but did not require) contextual metadata: geographic coordinates, units, min/max values, and timestamps. It was built to extend IFC (Industry Foundation Classes -- at the time a key data format for Building Information Modelling used in construction) to account for structures, devices and environments that move, change and adapt in real-time. The original specification was hosted at EEML.org.

In practice, Pachube also supported CSV (for constrained devices with limited processing power), JSON (for web developers), and Atom and RSS for interoperability with existing web tools. An EEML document describing a room's temperature sensor looked like this:

<eeml>
  <environment>
    <location exposure="indoor" domain="physical" disposition="fixed">
      <name>My Room</name>
      <lat>32.4</lat>
      <lon>22.7</lon>
    </location>
    <data id="0">
      <tag>temperature</tag>
      <value minValue="23.0" maxValue="48.0">36.2</value>
      <unit symbol="C" type="derivedSI">Celsius</unit>
    </data>
  </environment>
</eeml>

RESTful API

The main technical interface was a fully RESTful HTTP API -- meaning it used standard web verbs (GET, PUT, POST, DELETE) to read and write data, and could be interacted with using any HTTP client. The design decision to build on REST principles (not an obvious choice for connected devices at the time) rather than on proprietary protocols was deliberate because it meant that any developer already familiar with building RESTful web applications had essentially all the tools needed to connect a device to Pachube.

Authentication used API keys -- tokens sent with each HTTP request that identified the caller and determined their permissions. As the platform matured, CTO Ben Pirt designed a novel authentication system for massive scale, leading Pachube to introduce Advanced API Keys: a more granular permissioning system allowing keys to be scoped to specific feeds, specific datastreams, specific IP addresses, or specific referring URLs, and to be read-only or writeable. Ben's key innovation was a way of encoding some aspects of the scoping into the key so that invalid keys could be rapidly ignored by load balancers before passing permissions down the stack to more processing-intensive infrastructure -- something that became much more common later.

Advanced API Keys could be generated on the fly by devices themselves, enabling basic auto-provisioning. The November 2010 newsletter described this as "a new permissioning system for Pachube-connected devices that we think is going to be a major part of the 'internet of things': the ability to emphasise security, while still encouraging openness."

Triggers

Triggers were small functions attached to a datastream that watched for a condition -- a value exceeding a threshold, falling below one, or equalling a specific number -- and when that condition was met, made an HTTP GET request to a URL specified by the user. This meant Pachube data could drive external actions: sending an alert, updating another service, switching a device. In the period before cloud automation tools like IFTTT (which didn't launch until 2011), this was one of the few practical ways to connect sensor data to trigger actions at a wider or physical scale.

Technical Infrastructure

In 2010 the system ran on Ruby on Rails, with backend processes in Ruby, data stored in PostgreSQL, and traffic served by nginx load-balanced across a cluster of Thin processes via HAProxy. Memcached was used for in-memory caching. Hosting was on Rackspace Cloud, with redundant load balancers, a replicated database pair with hot standby, and nightly encrypted backups to Amazon S3. Later, the data store was migrated from PostgreSQL to a sharded Cassandra cluster, eventually enabling horizontal scalability across nodes in different datacentres.

Business Model

Pachube operated a freemium model. Basic access was free while premium accounts (including Pachube Pro, launched 02 June 2010) added full history access, privacy controls, SSL, higher API rate limits, statistical tools, and advanced trigger options. Pachube enterprise also provided bulk device registration, user management, white-label dashboards, and a device provisioning server, aimed at manufacturers wanting to offer Pachube-powered connectivity to their own customers.

Connected Environments was intended to serve as an integration services company, for city-scale and construction industry deployments but eventually disappeared into the background as its branding got confused with Pachube's.

Pachube's design and name changes through the years

Pachube went through several design iterations over the years. Shortly before being acquired we changed from 'pink' Pachube to 'yellow' Pachube, designed by Tom Williams. The dev team had always hated the map (it was difficult to maintain, presented unreliable information and took up valuable landing-page real estate) so it was removed in the new design. I disagreed somewhat, because I felt the implicit connection to the "real" world was important, and it was reintroduced a few months later once the site scalability issues were under control.

I had to wait to late 2013 to really get my full revenge though: launching Thingful.net -- a search engine for the internet of things with full screen map, satellite imagery and all.

---

5. Pachube's IoT App Ecosystem

Pachube.app IoT App Store

Pachube maintained an IoT App store at apps.pachube.com (archived here), a repository of tools that worked with data from any Pachube feed. Some were built by the Pachube team; others were contributed by the developer community:

• Graph Builder / PachubeViz -- zoomable, embeddable graphs of historical datastream data
• Dashboard -- a web-based control panel with a skeuomorph interface of switches and knobs for remotely switching actuators and adjusting dials, built on the Advanced API Key system
• Trails -- real-time maps showing the location of moving objects tracked through Pachube feeds
• Earth Browser -- a 3D globe interface for viewing datastream values distributed across the planet
• PachuBlog / PachuDial / Pachube Google Gadget -- embeddable widgets for blogs and websites
• SMS Alert -- monitoring tool that sent text messages when datastream conditions were triggered
• CarbonFootprint -- a real-time converter translating electricity datastreams into carbon footprint estimates, integrating with AMEE
• Porthole (Pachube AR) -- an augmented reality application using QR codes to display real-time sensor data overlaid on the physical objects or spaces that generated it
• Data Logger -- a generalised iPhone app for logging data to a Pachube feed with geolocation and timestamp metadata
• Pachube2Sketchup -- a real-time import of Pachube data into 3D SketchUp architectural models (see demo here)
• PachubeMon -- an iPhone feed monitoring application
• Feed Finder -- a search interface for discovering public Pachube feeds
• PachuME (J2ME) -- a Java Mobile Edition interface created by community member mchavez for basic cellphones

Pachube & Augmented Reality

We had announced a series of prototype augmented reality apps for mobile phones in early 2009, built by Chris Burman (who later left to tour the world as a pop star). These are documented on YouTube here and here and here.

Chris and I wrote a paper for W3C, Portholes & plumbing: how AR erases boundaries between 'physical' & 'virtual', to make the case that future AR standards should focus on facilitating communications between disparate realities rather than defining how, when or where they are experienced and that standards should be designed expressly to encourage lateral approaches in reality design.

The idea for integrating AR in Pachube was that if every feed had a location and a meaning, then you ought to be able to stand in front of the building, river or machine that produced the data and see its live readings overlaid on the thing itself. It was an attempt to close the loop between a sensor reading and the physical world it came from, coupled with people's ability to make in-situ immediate decisions based on what they saw.

---

6. Arduino, Physical Computing and the Maker Community

Shared Origins

The relationship between Pachube and the Arduino community was close from the beginning, based on shared networks and overlapping interests.

I had originally been introduced to Processing by Casey Reas during our residency at IDII in 2001, and later invited him to come talk to my students in the Bartlett's Interactive Architecture Workshop about how it might be used to connect screen-based interaction to the physical world, initially with Lantronix serial devices.

Once they became more readily available in London, Arduino, which launched publicly in 2005, was by 2008 becoming the dominant hardware platform for designers, hobbyist and artist physical interaction development. Initially connections to Pachube needed an internet connected computer or bridge to push data to the platform from Arduinos (typically using firmata).

At a workshop hosted by Duncan Wilson at London's Arup HQ in 2008, Massimo Banzi worked to get an early Arduino Ethernet shield connected to Pachube. In a sense Pachube was building the web server side of a similar problem that Arduino was addressing on the hardware side.

A formal agreement with Tinker London -- the London-based company, founded by Alexandra Deschamps-Sonsino, that was at the time the main European distributor and reseller of Arduino boards -- was signed in August 2008, establishing cooperation to link Pachube updates to Tinker's upcoming internet-enabled products. This gave Pachube an early formal connection to the Arduino ecosystem.

The Arduino community formed a significant portion of Pachube's early active user base, and Pachube provided the Arduino community with a place to store and shared devices' real-time data -- a public, persistent, queryable store for the sensor readings that would otherwise have been displayed only locally. The combination made it possible to do things neither platform could do alone: not just measure your home's energy consumption, but graph it over months, share it publicly, compare it with neighbours, or trigger alerts when it crossed a threshold.

In a broader sense, the Pachube-Arduino relationship was an early instance of a pattern that would later feature in much of the IoT industry where hardware platforms generated data and cloud platforms stored and processed it.

The PachubeClient Tutorial

The PachubeClient tutorial, created by Tom Igoe in March 2010, was published on the official Arduino website showing step by step how to connect an Arduino with an Ethernet shield to Pachube and upload a sensor reading every ten seconds. Adrian McEwen later authored the official Pachube library for Arduino Ethernet.

The tutorial became one of the most widely referenced starting points for anyone wanting to connect an Arduino to the internet -- which in that era often meant connecting it to Pachube. Its influence spread through the Arduino forum, Instructables, Make magazine, Adafruit's blog, and dozens of hobbyist project blogs, generating a long list of related projects and tutorials throughout 2010 and 2011.

Community Libraries and Projects

Pachube's community site and forum were very active from 2009 to 2012, hosting dozens of discussions, tutorials, projects and hardware/software recommendations.

I had created an EEML library for Processing -- enabling sketches to format and send data in EEML -- pushing data provided by Arduinos connected by firmata. Community members developed and maintained the Pachuino library (for Pure Arduino EEML output) and the ERxPachube library (for Arduino Ethernet shield integration).

Make magazine covered the Arduino-Pachube combination in June 2009, with a piece titled 'How-To: Web-connected sensors with Arduino + Pachube', featuring a community member who had connected two light sensors to an Arduino to compare indoor and outdoor light levels, forwarding the data to Pachube via Processing. The author described the process as "quite straightforward, making related projects much more approachable."

Adafruit Industries launched an Ethernet shield (as early as 2007 or 2008 if I recall) which eventually was one of the two options supported by the official PachubeClient tutorial. They featured Arduino-to-Pachube projects on their blog from 2010 onwards, helping spread the combination further into the electronics hobbyist community. Limor Fried / aka ladyada's Xport-based ethernet shield was initially the most useful one because it had DHCP capabilities that others' did not.

A significant milestone was when developers working with the Arduino Ethernet shield began achieving what one described as "a Pachube manual update client on the Arduino that pushes EEML XML format to Pachube directly, without a PC in the middle", eliminating the need for a computer running Processing as an intermediary, which had been the setup in many earlier Arduino-Pachube integrations. This made the combination substantially more practical for deployed, always-on monitoring applications.

Ed Borden had joined the team in early 2011 to help lead business development, which at the time primarily meant finding better ways to work with the community and helping make sure we stayed one step ahead as people developed novel uses for the platform. The diversity of what people actually built illustrates how broad the Arduino-Pachube overlap became. A selection drawn from forums, blogs, and the Arduino blog's Pachube category:

• How to build a web connected gas meter with your Arduino -- a London-based hobbyist built a web-connected gas meter using an Arduino, Ethernet shield, and a photoelectric sensor counting reflections on a dial, streaming consumption data to Pachube in EEML format. He had previously hacked a Current Cost electricity monitor and pushed that data to Pachube too.
• XBee wireless sensor networks -- hobbyists combined Arduino with XBee ZigBee radio modules to create multi-node wireless sensor networks, using one node as an internet-connected coordinator that forwarded all readings to Pachube. See A "Minimal" Arduino/XBee/Pachube Sensor Network
• Room temperature and light logging -- one of the commonest entry-level projects; hundreds of people's first Arduino-to-Pachube feeds measured temperature or light levels in a room
• Solar water heating systems -- monitoring of home solar installations, comparing actual and predicted performance over time. See Easter Hack - An Arduino Solar Thermal Controller
• Elevator usage tracking -- an architect used Pachube to record real-time elevator usage data from an academic building. See interactive elevator installation
• Air quality and radioactivity -- some users were building multi-sensor boxes combining air quality measurement with Geiger counter inputs before the Fukushima crisis brought that combination into sharp focus
• Air Conditioner Web Remote Control with Arudino

Nanode

Another connection to Home Camp was Ken Boak who had created the Nanode, essentially an Arduino with built-in ethernet, which became popular for hobbyists and designers working on networked physical interaction. For a period they shipped with one year of Pachube Premium service free.

mbed

Alongside Arduino, the mbed rapid microcontroller prototyping platform also developed a strong community of Pachube users, with tutorials and libraries available at mbed.org/cookbook/Pachube. The November 2010 newsletter noted "growing popularity in using Pachube and mbed, a tool for rapid prototyping with microcontrollers, now that a number of examples and tutorials are available."

Hackaday

Hackaday had several articles about using Pachube, including How To Add Flair To A Webpage That Shows Your Project Data, Monitoring Air Quality With A $12 Sensor, Detecting Seismic Waves with a Piezo Element and Data Scraping And Visualization With Python.

---

7. Academic and Technical Influence

Textbook References

The most detailed published technical analysis of Pachube appeared in Interconnecting Smart Objects with IP: The Next Internet by Jean-Philippe Vasseur and Adam Dunkels (Morgan Kaufmann/Elsevier, 2010). Chapter 9, Section 3 -- "Pachube: A Web Service System for Smart Objects" -- used the platform as the worked example for RESTful interface design for smart objects.

Pachube appeared in dozens of other books. See in particular IoT Mashups with the WoTKit, by Michael Blackstock and Rodger Lea; and "Getting Started with the Internet of Things" (Cuno Pfister, O'Reilly, 2011), which contains a dedicated Chapter 5: "Pachube" and Chapter 6: "Hello Pachube."

Pachube was also featured extensively in Arduino Projects to Save the World by Emery Premeaux with Brian Evans.

Industry Research

McKinsey Quarterly's 2010 article "The Internet of Things" cited Pachube in its analysis of how networked sensors would transform industries. Harbor Research's Emergence of Smart Business (2010) listed Pachube as one of seven case studies in its M2M and Smart Systems market forecast.

The architectural patterns Pachube established -- RESTful HTTP APIs for sensor data, hierarchical feed/datastream data models, API key permissioning, trigger-based callbacks, app ecosystems layered over open data -- became standard features of IoT platforms that followed. ThingSpeak, Carriots, Open.Sen.se, and others that launched in subsequent years operated within a conceptual framework that Pachube had helped define. The separation of data collection from data application -- Pachube as infrastructure, with applications built on top -- became a recurring model for the IoT platform industry.

---

8. Key Projects and Partnerships

Current Cost and the Bridge

In June 2010, at the second annual Internet of Things conference in Brussels, Pachube announced a partnership with Current Cost -- a UK manufacturer of home energy monitors with over 1.8 million units deployed at the time. Current Cost used Pachube to power the Bridge, an ethernet device that connected Current Cost electricity monitors to the internet, enabling real-time energy use analysis via web dashboards, smartphones and a full array of Pachube apps.

The Bridge (designed by Tinker London) included enterprise-level features -- privacy groups, a statistics API, user management, and a device provisioning server -- while remaining accessible to consumers, who could claim their Pachube feeds to make them public or otherwise use their data as they wanted. Current Cost CEO Martin Dix said at the time: "We believe in connected devices, and Pachube allowed us to build a connected service faster than we could have built on our own. The next step is to empower our users to share their data and collaborate with other users to make better decisions." - Pachube press release, "Current Cost's 'Bridge' uses Pachube for backend data management", 03 June 2010.

Cisco Urban Ecomap 2.0

In 2010, Cisco used Pachube as the data infrastructure for Urban Ecomap 2.0, a tool for tracking and visualising building-by-building carbon emissions across cities, part of Cisco's Connected Urban Development initiative, with JD Stanley, Shane Mitchell and Steve Tiell.

Atos Origin

Pachube also collaborated through 2010–2011 with Atos Origin (renamed Atos in 2011): Atos built a technical proof-of-concept connecting Linky, EDF/ERDF's French smart-meter rollout, to Pachube, and I presented our work to Atos's internal "Scientific Community" innovation group in Paris at the invitation of Hubert Tardieu, then a senior advisor to the Atos CEO. Working with Guy Lidbetter (Atos CTO, Managed Operations) and Mike Smith (Atos Chief Architect), we also carried joint architecture and business-model reviews including with Atos's Barcelona subsidiary Tempos 21.

Pachube and IPv6

On June 7, 2011, Pachube announced a collaboration with Worldsensing, Cisco, and AT&T to Support IPv6 Internet of Things.

Twilio & Chumby

Competitions we announced included Getting Physical with Twilio and Quick!! Build an awesome Pachube-Chumby app and win!.

W3C Points of Interest Working Group & Augmented Reality

Discussion about Pachube at W3C, including a meeting and an online discussion, hosted by Christine Perey.

See also W3C Augmented Reality Workshop — "Portholes and Plumbing", paper by Burman & Haque available here; and W3C public-lod list — "EEML: Extended Environments Markup Language": linked-data researchers discussing how to represent EEML/Pachube feeds as RDF, referencing the EEML 0.5.1 schema, a connection to the semantic-web community.

---

9. Milestones and Community Events

Pachube Global IoT Hackathon, April 2011

On 8-9 April 2011, coinciding with Global Internet of Things Day, Pachube ran a 24-hour Global Internet of Things Hackathon. Starting at 2pm UK time, the event brought together developers, designers, and makers at events in London (the main venue), New York City, Tokyo, Amsterdam, Zurich, Eindhoven, Linz, Lancaster and other cities.

Participants all around the world worked on IoT projects using each other's live Pachube data feeds. Among other presentations at the London node, Nick Weldin introduced Arduino and Ken Boak talked about Nanode.

See Rain Rabbit's Flickr album for lots of photos.

Ben Bashford, Tim Burrell Saward and Dan Williams created an augmented-reality Display Cabinet at the event. Here is some info from Hackteria's event node held in Zurich documenting a prototype of a turbidity sensor shield.

The Internet of Things Bill of Rights, March 2011

In early 2011, Ed Borden and I had become particularly concerned about data usage rights, particularly in the context of wanting to make it possible to create an open, not closed, internet of things. We drafted a Pachube Internet of Things Bill of Rights for Pachube's community site, and published it later that year in Volume magazine under the heading "Data and Owner." It was one of the first systematic attempts to articulate data rights in the context of connected physical devices:

1. People own the data they (or their 'things') create.
2. People own the data someone else creates about them.
3. People have the right to access data gathered from public space.
4. People have the right to access their data in full resolution in real time.
5. People have the right to delete or back up their data.

The document was subsequently taken up by the Open IoT Assembly, held in London in 2012, which brought together 85 participants to expand and ratify a broader statement of open IoT principles. Speakers at the event, organised by Alexandra Deschamps-Sonsino and myself, included Adam Greenfield, Rob van Kranenburg.

Pachube in Museum of Modern Art "Talk to Me" exhibition

Curated by Paola Antonelli, Pachube was included in MoMA's "Talk to Me" exhibition as a designed system representative of the Internet of Things.

London Internet of Things Meetup / IoT London Meetup

The IoT London Meetup grew out of Pachube's local developer community, curated and organised by Alexandra Deschamps-Sonsino from autumn 2011 onwards, using a familiar all-caps yellow+orange identity that became a standard for other international IoT meetups.

Over the next decade, 100 meetups were held in London, drawing new and returning attendees from across the city and beyond, with speakers from around the world. Hosted at various venues across the city, including Pachube's own offices early on, the meetup became, through Alex's tireless energy, the second largest dedicated IoT meetup on Meetup.com globally.

---

10. Fukushima and Citizen Science Radiation Monitoring

On 11 March 2011, the Tōhoku earthquake and tsunami triggered meltdowns at the Fukushima Daiichi nuclear power plant. In the days that followed, Japan's official radiation monitoring systems were damaged, and the government was slow to release data to the public. The SPEEDI dose prediction system generated data, shared with US military forces but not released to the Japanese public until two weeks after the disaster. Where data was available it was frequently out of date and only distributed in PDF format.

We quickly noticed that Pachube users had started creating radiation feeds using their own individual geiger counters and radiation monitoring equipment, so we gave free upgrades to accounts of all users based in Japan, unlocking premium features, including high resolution history. Within days, members of the Pachube community in Japan had connected dozens of geiger counters and monitors to the internet, streaming readings to Pachube feeds. The platform became a central aggregation point for what grew to hundreds of independent radiation feeds from across Japan.

IEEE Spectrum covered the effort in a piece titled "Radiation Monitoring in Japan Goes DIY" describing how "a group of device hackers are digitizing Geiger counter readings to monitor radiation levels in real time across Japan. The data is being aggregated by the platform Pachube."

There was criticism in the media at the time of this "citizen science" effort, particularly because there was little uniformity in the users' setups, equipment or even data units.

These critiques tended to be formed round the assumption that the goal behind citizens' efforts was data-collection, aiming to amass a large quantity of data in order to conduct 'objective' evaluation of it. From this perspective, 'messy' data was inherently not useful and only uniform, calibrated data could be meaningful.

To me, this placed far too much emphasis on the data itself (I used to talk at the time of society of the data spectacle) and disregarded the key motivation that I actually heard from community users: that the act of measuring and sharing data was simply a way to do something proactive, to be part of a public conversation round a rapidly-changing situation, and to make very visible the idea that people could learn from their own environment and take action themselves by learning from others, rather than just waiting for things to be done on their behalf.

They were not, as far as I could see, particularly interested in contributing to a scientific process of broad data aggregation, and the primary sense-making came through being able to have conversations with other people because of declared shared interest, not from the data itself -- so the distinction between 'messy' and 'clean' data didn't seem particularly relevant.

The Pachube community started building a range of applications using Pachube to visualise and make sense of the real-time data (and its relationship to other data on the platform like wind). One developer published a widely shared tutorial on converting analogue Geiger counter "chirps" into digital data uploadable to the web. The resulting feeds were used both directly by individuals in affected areas and as data sources for independent visualisation and alerting tools built by developers elsewhere. Wikipedia's article on radiation monitoring in Japan documents the effort, and Hannah Wolfe wrote about it as well.

The initiative was also covered several times by Make magazine, including DIY Geiger Counters Take Center Stage and Hackerspace Happenings: MAKE Interviews Tokyo's Akiba

---

11. The Air Quality Egg

Returning from a trip to China in spring 2011, I mentioned to Ed Borden that air quality monitoring seemed to be a major citizen concern with destination web pages even embedding real-time data feeds from AQ monitors. We realised that this was similar to the way that energy monitoring had recently entered cultural consciousness in Europe, and radiation in Japan. We noticed also that the number of air quality feeds on Pachube had been growing steadily throughout late 2009 and 2010. With this in mind we decided to host a series of air quality monitoring workshops, sponsored by Pachube.

From one of these was born the Air Quality Egg (AQE), a low-cost, open-source, networked air quality sensor measuring nitrogen dioxide and carbon monoxide, created for members of the public to deploy at home and at work.

The device was designed to stream data to Pachube, which served as the shared public dataset aggregating readings from deployed eggs. The AQE went on to raise funding through Kickstarter in 2012, where it was named one of the best projects on the platform. Fast Company's The Air Quality Egg Will Let You Know Exactly What You're Breathing wrote that it would "turn everyone into an atmospheric scientist (or at least a data collector)". In the early years between 1,000 and 2,000 units were manufactured by WickedDevice and deployed internationally (described further here by Citizen Sense research project at Cambridge and here) though later went in a more commercial direction.

---

12. Press Coverage

Pachube attracted a lot of press coverage from 2008 onwards. A selection from the early days:

• CNN.com digitalBiz -- 'Internetting Every Thing, Everywhere, All the Time', November 2008
• Spime! Internetting every thing, everywhere, all the time" — Make by Phillip Torrone
• The Guardian -- 'The net shapes up to get physical', 16 October 2008
• Le Monde / InternetActu -- 'L'internet des objets va-t-il changer la nature des objets?', June 2009
• The Economist, Technology Quarterly -- 'Sensors and Sensitivity', June 2009: described Pachube as "a first glimpse of what such a network would look like"
• New York Times / ReadWriteWeb -- 'Top 10 Internet of Things Products of 2009', December 2009: Pachube listed first
• Guardian DataBlog -- 'How to Friday: Pachube, Sensors & Augmented Reality', October 2009
• Frontier Economy -- 'Wiring Up the Internet of Things', October 2009
• ReadWriteWeb -- 'Pachube: Building a Platform for Internet-Enabled Environments', September 2009
• The Economist, Special Report on Smart Systems -- 'Augmented Business', November 2010
• ReadWriteWeb -- 'Pachube Partners with Current Cost', June 2010
• Information Age -- 'Pachube opens the Internet of things to end users', 2011
• iM2M (Chinese M2M magazine) -- 'Smart Things, Living Environments', November 2010 (print only, no URL)
• CIO Insight (China) -- 'Internet of Things', January 2010
• MIT Technology Review, April 2011
• NZ Herald -- An Invitation to Write the Future, 7 April 2011
• Several other articles in the Guardian in 2011: Creative Data: Technology and Supply Chain Transparency, Smart Cities and Clean Technology, The Internet of Local Government Things

---

13. Acquisition History

LogMeIn, July 2011

It had been very hard to raise investment for Pachube (despite all the industry enthusiasm I was declined by around 170 VCs and institutional investors) and as our costs rose it became essential to plan for the platform's future.

In July 2010 an email had arrived in my spam folder with the forgettable subject line: "what is your address and phone number?" I was about to delete it, but when I opened the email it had no content but a LogMeIn signature -- which seemed odd for a corporate sender, neither spam nor phishing -- so I replied and a conversation started.

Almost a year later, on 19 July 2011, after months of discussion and negotiation, LogMeIn acquired Pachube (note that this press release and others that reference it vastly overstate its running costs! I prided myself on running a lean company...) The Pachube team, including me, would remain with the platform. LogMeIn framed the acquisition as an extension of its cloud connectivity services beyond computers and smartphones to internet-connected devices more broadly.

Under LogMeIn, we worked to rebuild and scale the platform within their distributed infrastructure. The service was first rebranded as Cosm (briefly, in 2012, retracted due to a trademark issue), then relaunched as Xively in May 2013, positioning itself as a "Public Cloud for the IoT" aimed at enterprise customers rather than the original hobbyist and developer community.

The technical architecture was extended, adding features around identity management, device provisioning at scale and enterprise security. Xively won the 2014 GSMA Mobile World Congress Best Cloud-Based Technology for Mobile, and was named one of Fast Company's World's Top 10 Most Innovative Companies in the Internet of Things in 2014.

Google, 2018

In February 2018, Google announced its intention to acquire Xively from LogMeIn, an acquisition that closed in March 2018. LogMeIn had by then decided to refocus on its core customer service and identity access businesses. Google's stated rationale was to use Xively's device management, messaging, and dashboard capabilities to strengthen Google Cloud IoT Core, its managed IoT service. All 45 Xively employees transferred to Google, relocating to Google's Cambridge offices. TechCrunch, 'Google to acquire Xively IoT platform from LogMeIn for $50M', 16 February 2018; Google Cloud Blog, 15 February 2018.

Google Cloud IoT Core was itself subsequently deprecated in 2023, along with all Pachube-related domains, ending the direct lineage that had begun with the first EnvironmentXML prototype in 2004.

---

14. Pachube's Contribution to the IoT landscape

Pachube operated at a time when the idea of the Internet of Things was shifting from a research concept to something people were actively trying to build. When it launched publicly in 2008, there were no cloud IoT platforms, industry-standard device connectivity protocols in wide use or shared vocabularies for what a "sensor feed" or a "connected environment" should look like on the web.

What we, the Pachube community as well as the Pachube team, arrived at through practical experimentation -- data organised into feeds and datastreams, HTTP-based APIs for reading and writing, per-device API keys for access control, app ecosystems layered over shared open data -- became, over time, the common grammar of the IoT platform industry.

This influence operated through several channels, initially through the community of developers who experienced features on Pachube, requested new ones, and carried the practice into their subsequent work. Pachube was also widely published in academic and technical literature, with dozens of chapters in textbooks and technology overviews.

The major cloud IoT platforms that followed (e.g. AWS IoT (2015), Azure IoT Hub (2014), Google Cloud IoT Core (2017)) were created by companies with vastly more resources and operated at a much larger scale. They built on a set of architectural conventions and community expectations that earlier platforms like Pachube had done much to establish.

The specific design decisions that we made with Pachube -- RESTful HTTP APIs for sensor data, hierarchical feed/datastream data models, API key permissioning, trigger-based callbacks, device-agnostic data ingestion, app ecosystems layered on top of open data -- became standard features of subsequent IoT platforms.

The IoT Bill of Rights, and the Open IoT Assembly it fed into, were early contributions to an ongoing set of debates about IoT data, surveillance, data sovereignty, and the rights of people in relation to the devices around them. The principles articulated in 2011 -- ownership, access, portability, deletion -- remain contested and unevenly realised in practice.

While Pachube was part of what seemed to be a trajectory towards openness, this unfortunately never fully materialised in the industry (or indeed in the broader technology landscape.) We now see, instead, thousands of services and platforms living (and dying) in isolated "walled gardens", with very few common standards to connect them, and even fewer possibilities to build the internet of things as common, shared public good. I've written about that in How Might We Grow Diverse Internets of Things? Learning from Project Xanadu & the WWW.

15. What came next

After the acquisition by LogMeIn, where I spent two years leading the team and then heading up the Urban Projects Division, I went on to launch Umbrellium a design studio for urban technology and communities in 2013; Thingful, later that year, a search engine for IoT data, designed to make public sensor feeds discoverable across a wide variety of emerging IoT platforms; Starling Technologies, infrastructure for interactive roads and crossings in (co-founded in 2020, with Dot Samsen and Andrew Caleya Chetty); and HAQUE TAN, a design studio (with Ling Tan) combining the scale of architecture with the ingenuity of art and the eccentricities of technology, to make spaces more democratic, inclusive and culturally‑driven.

---

16. Timeline

Year	Event
Jun 1996	My postgrad project 'Conversations with a kettle, and other friendly household objects', at the Bartlett School of Architecture.
2004	Presented "RSS feeds for environments" concept at the RAM5 conference in Riga, Latvia; began development of EnvironmentXML (EXML) prototype
12 Feb 2005	EnvironmentXML (EXML) launched publicly, a platform for connecting and sharing real-time data between physical environments
2005	Collaborated with Adam Somlai-Fischer on Low tech sensors and actuators for artists and architects
2006	EnvironmentXML grows to around 70 users by mid-year; new registrations disabled due to server costs.
2007	Connected Environments Ltd incorporated; Sam Mulube begins building the platform in October. In December 2007, platform renamed from 'Connected Environments' to Pachube
28 Dec 2007	Pachube.com domain registration
19 Feb 2008	Pachube launched, invitation-only
Jun 2008	Workshop at Arup's London headquarters, hosted by Duncan Wilson, with Massimo Banzi
Aug 2008	Agreement with Tinker London to cooperate on Pachube-enabled products
Nov 2008	CNN covers Pachube in 'Internetting Every Thing, Everywhere, All the Time'
Jan 2009	Tish Shute "Pachube, Patching the Planet" interview published -- a detailed articulation of the platform's aims and its relationship to my design practice
Feb 2009	Augmented reality apps for mobile phones announced, built by Chris Burman, overlaying live Pachube feed data onto the physical objects and spaces that generated it
Apr 2009	Workshop at PixelAche festival; Pachube community member publishes tutorial on building a web-connected gas meter with Arduino and Pachube
Jun 2009	The Economist Technology Quarterly covers Pachube. Le Monde covers Pachube in a piece on the Internet of Things
Jun 2009	Make magazine publishes 'How-To: Web-connected sensors with Arduino + Pachube'
Aug 2009	Pre-seed investment round completed, enabling the platform to support and scale as open IoT infrastructure
Oct 2009	Guardian DataBlog: 'How to Friday: Pachube, Sensors & Augmented Reality'
Dec 2009	Ben Pirt joins as CTO. New York Times / ReadWriteWeb: Pachube named first in 'Top 10 Internet of Things Products of 2009'
Jan 2010	Pachube nominated for Design Museum Design of the Year (Interactive)
21 Jan 2010	Platform opened for public signup
Mar 2010	Tom Igoe (Arduino co-founder, ITP New York) publishes PachubeClient tutorial on arduino.cc
02 Jun 2010	Pachube launches 'premium' Pachube Pro and Pachube Enterprise plans and starts expanding a revenue model
Jun 2010	Pachube-Current Cost partnership announced; Current Cost Bridge launched, powered by Pachube
2010	Cisco uses Pachube as data infrastructure for Urban Ecomap 2.0
2010	Collaboration with Atos Origin on smart-grid/M2M integration; presentations to Atos's Scientific Community innovation group
2010	Interconnecting Smart Objects with IP (Vasseur & Dunkels, Morgan Kaufmann) publishes full technical analysis of Pachube as canonical RESTful IoT API
Nov 2010	Pachube in The Economist Special Report on Smart Systems: 'Augmented Business'
Nov 2010	Pachube cited in Harbor Research's Emergence of Smart Business forecast report as one of seven M2M and Smart Systems case studies
Nov 2010	Advanced API Keys feature launched
Mar 2011	Pachube IoT Bill of Rights published in Volume magazine
11 Mar 2011	Tōhoku earthquake and Fukushima nuclear disaster. Pachube community members begin aggregating Geiger counter feeds from across Japan and adding hundreds of radiation datastreams; also building apps to visualise radiation data
Apr 2011	IEEE Spectrum: 'Radiation Monitoring in Japan Goes DIY' -- Pachube coverage
8-9 Apr 2011	Pachube's 24-hour Global IoT Hackathon, with events in London, New York, Tokyo, Amsterdam, Zurich, Eindhoven, Linz, Lancaster and others
7 Jun 2011	Pachube announces collaboration with Worldsensing, Cisco, and AT&T to support IPv6 Internet of Things
19 Jul 2011	Pachube acquired by LogMeIn
24 Jul 2011	Pachube included in MoMA's Talk to Me exhibition
Sep 2011	IoT London Meetup founded, growing out of the Pachube developer community, organised by Alexandra Deschamps-Sonsino
2012	Air Quality Egg Kickstarter campaign, growing out of air quality monitoring workshops hosted by Pachube
May 2012	Service briefly rebranded as Cosm
17 Jun 2012	Open IoT Assembly, London -- expands the Pachube IoT Bill of Rights into broader principles
May 2013	Relaunched as Xively, targeting enterprise customers
Sep 2013	I launched Umbrellium (a design studio for urban technology and communities) and later in the year Thingful (a search engine for IoT data)
2014	Xively wins GSMA Mobile World Congress award for Best Cloud-Based Technology for Mobile; named in Fast Company's Top 10 Most Innovative IoT Companies
Feb 2018	Google acquires Xively from LogMeIn; technology absorbed into Google Cloud IoT Core
16 Aug 2023	Google Cloud IoT Core retired, ending the Pachube lineage