All posts tagged: things

One of the major hurdles of developing portable (and connected) devices is finding the balance between power consumption and battery storage that allows for a genuinely useful device and experience. Generally most components can be optimised through good design and wise choices, however the main microcontroller or CPU can be a sticking point – until now.

Intel have taken this problem to heart and as a solution, recently announced their “Quark” family of system-on-chip cores. They’re a family of low-power 32-bit CPU cores designed to compete with ARM’s Cortex-M series in modern Internet-of-Things and wearable embedded computing applications.

Quark is a very low-power and compact x86-compatible core designed to be even smaller and lower in power consumption than Intel’s low-power Atom CPU cores, which are targeted at tablets, low-power netbooks and smartphones.

Notably, Quark is the first Intel core that is fully synthesizable and designed for potential integration with third-party IP blocks. This means that a customer could use the Quark core, license it from Intel, and hook it to peripherals on a custom system-on-chip, like for example custom graphics, I/O, storage, 802.11 or 3G networking.

It is claimed that Quark will be one-fifth of the size of the Atom core, and have one-tenth of the power consumption. At this level, Quark is much more powerful – and power hungry – than a lightweight 8-bit microcontroller, but it is also not a competitor to the more powerful ARM Cortex-A family either. It aims to compete with the popular Cortex-M family of 32-bit microcontroller cores from synthesizable microcontroller IP leader ARM.

The Quark core is a single-core, single-thread, low-power, small-footprint CPU core, and it is targeted at “Internet-of-Things” applications, wearable computing devices such as “smart watches”, and low-cost disposable medical devices as well as industrial and building automation control systems.

At this years’s Intel Developers Forum, a prototype “smartwatch” based on Quark technology was displayed as a proof of concept, along with a wearable instrumented patch for medical datalogging. Quark has been demonstrated in a prototype Internet-of-Things enabled HVAC automation application by HVAC leader Daikin. Daikin’s prototype system has WiFi and 3G support, and allows for secure remote control and monitoring.

The Quark product line is designed to slot in below the existing Atom family in terms of cost and power consumption, compatible with the Pentium instruction set architecture but aimed at markets where small form factor and low power consumption take priority, with a power consumption target that is apparently less than 100 milliwatts in some cases.

This power efficiency makes Quark attractive in wearable computing applications such as “smart watches” and Google Glass style wearable displays where battery capacity is very limited due to size constraints. Some bracelet-like wearable devices have been shown at this year’s Intel Developers Forum as a proof-of-concept of a wearable system powered by Quark technology.

Being smaller, lower power, and less powerful than Atom, Intel will be targeting the Quark product line at the Internet-of-Things market in applications where more power than a traditional embedded microcontroller is desirable or required, but less power consumption than an ordinary PC or notebook is desirable.

Quark is synthesizable, which means that customers can add their own IP around the core. ARM, for example, lets companies license its CPU cores and then add their own co-processors or other components to create chips optimised for a wide variety of projects and industries. How this would work in the case of Quark is not exactly clear however, since Intel plans to keep manufacturing of Quark silicon entirely in-house, at least initially.

This is a new move for Intel, but the company intends to retain control over their entire chip fabrication process in-house, bringing in existing customer IP for integration with Quark and in-house fab, although it is possible at least in principle that other foundries could fabricate Quark-based systems for licensees of the IP.

Intel’s decision with Quark means leveraging its own IP in a way that lets it offer customisable hardware to potential customers, without giving up control of either its processor IP or its own fab capabilities. Designers will not be allowed to customise the Quark core, they can only connect third-party IP blocks to its fabric.

Quark’s partially-open fabric appears to be somewhat derivative of ARM’s long-standing and successful policy of licensing its Cortex IP to other chip makers in a synthesizable form. ARM Cortex M3 and M4 cores have been rapidly stealing market share away from other microcontroller platforms in recent years – since the 32-bit architecture offers significant performance gains over 8-bit platforms such as PIC or AVR.

Furthermore their Cortex-M3 is finding its way into smartwatches such as the Sony SmartWatch 2 and the Qualcomm Toq as well as wireless sensor network system-on-chips such as TI’s CC2538 802.15.4/ZigBee/6LoWPAN platform. However as the Quark matures we’re sure it will be a successful player in the portable device and IoT arena.

Technologies such as Intel’s Quark are an example of how technology is constantly improving, and with the right knowledge it can be used to your advantage. However there are also many existing power-saving chipsets on the market your team may not be aware of, or unsure about taking on a new development platform.

But don’t let that get in the way of improving your existing or new designs – if you’re not sure about your options, discuss them with a team that understands the latest technologies, platforms and how to integrate them for your advantage – the team at the LX Group.

Getting started is simple – for a confidential discussion about your ideas and how we can help bring them to life – click here to contact us, or telephone 1800 810 124.

LX is an award-winning electronics design company based in Sydney, Australia. LX services include full turnkey design, electronics, hardware, software and firmware design. LX specialises in embedded systems and wireless technologies design. https://lx-group.com.au

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

Today’s homes are becoming increasingly connected and “smart”, and previously proprietary systems for applications such as energy monitoring, security and home automation are now moving towards increasing compliance with standard, open protocols such as IP to provide interoperability of the network and connectivity out to the Internet.

One interesting system known as Sensinode – part of the ARM organisation – offers a complete software solution for connected home applications, providing end-to-end software products that bring IP connectivity and web services right out to the end nodes in wireless Internet-of-Things networks, combining highly optimised embedded client software with a scalable management and web application platform.

Sensinode’s NanoStack, NanoRouter and NanoService solutions are valuable building blocks for the Internet of Things – mainly targeting large-scale mesh networks that require IP backbone connectivity such as home and building automation systems, sensor mesh networks for industrial control and monitoring, meter-reading systems and “smart” street lighting systems.

For example, Sensinode’s Connected Home Reference App provides a great starting point to rapidly deploy a web service for a connected home application, including a complete graphical web-based application for home automation and monitoring with floor plan import, device monitoring and control, data graphing and configurable notifications and alarms – an ideal starting point for home automation developers.

The intelligent control and monitoring of lighting systems is another area with great potential for reduced energy consumption and reduced maintenance costs through intelligent monitoring and control. NanoStack and NanoRouter provide low-power wireless IP connectivity for radio platforms ideally suited for indoor and outdoor lighting, while NanoService provides an end-to-end solution for the integration of lighting control and monitoring with web services.

Sensinode’s Street Lighting Reference App provides OEMs and system integrators with the tools to rapidly deploy a web-based service. Leveraging the power of the NanoService platform, this Street Lighting app includes Google Maps integration with real-time light monitoring and control, alarms, firmware updates and light group management.

Using Sensinode’s reference apps, complete graphical web applications and example source code provided, developers are able to easily get started developing and deploying machine-to-machine services on Sensinode’s NanoService platform. NanoService provides efficient embedded end-to-end web-service connectivity, integrating with new and existing backend web services.

NanoStack is Sensinode’s advanced 6LoWPAN protocol stack product for 2.4 GHz and sub-gigahertz 802.15.4 RF chipsets, while Sensinode’s NanoRouter software acts as a 6LoWPAN edge router, enabling routing between 6LoWPAN and IPv4/IPv6 networks. NanoRouter integrates with home and mobile 802.15.4 gateways and 802.15.4-enabled devices such as smart meters, providing Internet routing to the 802.15.4/6LoWPAN network.

The NanoStack communications stack for IP-based wireless sensor networks is platform- and radio-independent and gives hardware manufacturers, OEMs and system integrators a fast, easy and cost-effective way to harness Sensinode’s 6LoWPAN low-power mesh technology on inexpensive RF chipsets and microcontroller radio system-on-chips.

As NanoStack features support for both 2.4 GHz and sub-gigahertz 802.15.4 transceivers, robust performance and the ability to support consumer applications on the same system-on-chip as the IP network stack. NanoStack 2.0 fits the power of 6LoWPAN into a footprint of only about 10 kilobytes of flash memory in a low-cost embedded device. Entire wireless sensor firmwares are possible, using NanoStack 2.0, in a footprint of less than 32 kB of flash and 4 kB of RAM.

As NanoStack is designed to run on cheap, low-power, resource-constrained microcontrollers with embedded RF transceivers, such as the Texas Instruments CC2530 and CC430 system-on-chip devices, to name just a couple of examples. Rather than providing any hardware solutions themselves, Sensinode’s products are just software solutions that are used in conjunction with hardware platforms from third-party hardware providers such as Texas Instruments and Atmel.

Atmel has licensed Sensinode’s 6LoWPAN software stack for use with their ultra-low-power wireless hardware platforms. Sensinode’s NanoMesh and NanoService solutions are available on the Atmel Gallery for download, providing developers using Atmel hardware with a valuable starting point for the development of Internet-of-Things solutions using Sensinode technology.

As an example of a combined hardware and software solution implementing Sensinode technology, the Texas Instruments CC1180 6LoWPAN Network Processor is TI’s CC1110F32 low-power sub-gigahertz RF system-on-chip IC pre-loaded with Sensinode’s NanoStack 2.0 Lite 6LoWPAN stack.

The CC1180 handles all the timing-critical and processing-intensive 6LoWPAN protocol tasks in your Internet-of-Things application, leaving the resources of the application microcontroller free to handle the application. The CC1180 makes it easy to add 6LoWPAN functionality to new or existing products, as it provides great flexibility in the choice of the application microcontroller.

The CC1180 IC comes pre-loaded with a bootloader, Sensinode NanoBoot. This bootloader is used to download the Sensinode 6LoWPAN stack, NanoStack 2.0 Lite. The CC1180 offers simple integration of 6LoWPAN with mesh support into any design, running Sensinode’s mature and stable 6LoWPAN mesh stack.

This platform offers a simple UART interface to the host microcontroller, and the 6LoWPAN stack can be updated using the Sensinode NanoBoot API. Over-The-Air firmware updates are supported, provided that the host microcontroller has enough memory to store the new stack image.

With our existing experience in producing a wide range of devices incorporating embedded wireless technology our engineers can take your ideas for home or other types of automation to the final product stage using Sensinode or almost any other platform.

We can create or tailor just about anything from a wireless temperature sensor to a complete Internet-enabled system for you – within your required time-frame and your budget. For more information or a confidential discussion about your ideas and how we can help bring them to life – click here to contact us, or telephone 1800 810 124.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

Home automation is an emerging field with great potential, however without the appropriate standardisation of devices it can become a minefield of incompatibilities and frustrated customers. However there’s a standard we’re excited about – Zigbee Home Automation – that is quite promising.

ZigBee Home Automation is an application profile for Networked devices for home automation use – a global standard helping to create smarter homes that enhance comfort, convenience, security and energy management in the home environment. This standard for ZigBee wireless mesh-networked home automation applications can help make every home a smarter, safer and more energy efficient environment for consumers and families.

The standard gives your customers a way to gain greater control of the functionality of their home. By offering a global standard for interoperable products you it enables the secure and reliable monitoring and control of technologies in the home environment with robust, energy-efficient and easy to install wireless networks. Almost anything can be connected, such as appliances, home entertainment, environmental control and sensing, HVAC and security systems – providing convenience and energy efficiency benefits to the resident.

Smarter homes allow consumers to save money, be more environmentally aware, feel more secure and enjoy a variety of conveniences that make homes easier and less expensive to maintain. ZigBee Home Automation supports the needs of a diverse global ecosystem of stakeholders including home owners or tenants, product manufacturers, designers and architects, offering a standard that provides a reliable, consistent way to wirelessly monitor, control and automate household appliances and technologies to create innovative, functional and liveable home environments.

Typical application areas for ZigBee Home Automation can include smart lighting, access control, temperature and environmental sensing and control, intruder detection, smoke or fire detection, automated occupancy sensing and automated lighting or appliance control. The use of wireless radio networks eliminates the cost and effort of cable installation throughout the home, whilst the ZigBee standard provides certified interoperability and global 2.4 GHz ISM spectrum allocation, allowing manufacturers to take their ZigBee-based solutions to the global market relatively easily with relatively simple installation and operation.

Devices will have a typical RF range of up to 70 meters indoors or 400 meters outdoors, offering a flexibility to cover homes of all sizes. As with all ZigBee solutions, ZigBee Home Automation systems are built on top of an open and freely available specification based on international standards and represent a highly scalable solution with the ability to potentially network thousands of devices.

The devices are easy to install, even allowing for do-it-yourself installation in most cases. Employing wireless radio networks as well as battery power in many cases means that ZigBee Home Automation devices require little or no cable installation, making them ideal for easy retrofitting to existing homes and buildings as well as remodelling and new construction. Self-organising networks with easy device discovery simplify the setup and maintenance of networks consisting of many nodes, and the proven interference avoidance mechanisms in ZigBee networks ensure worry-free operation even in environments where coexistence with other 2.4 GHz radios such as 802.11 WiFi and Bluetooth is required.

The ZigBee Home Automation standard is designed for full coexistence with 2.4 GHz IEEE 802.11 wireless LANs and Bluetooth, as with all ZigBee technologies. Thus all devices based on these standards are designed to operate effectively in the same environment as WiFi networks, employing proven interference avoidance techniques such as channel agility.

Internet connectivity to the ZigBee network allows ZigBee Home Automation devices to be controlled via the Internet from anywhere in the world, as well as allowing WiFi-connected smartphones to be used as compact, powerful control and user-interface appliances to control the network of ZigBee appliances around the home.

Furthermore the standard is secure – employing AES128 encryption and device authentication to secure personal information, prevent unauthorised control of or access to the network, and to prevent interference or unauthorised access between independent neighbouring networks.

ZigBee Home Automation devices can be used to monitor household energy use, and to turn on and off devices remotely. Since ZigBee Home Automation is a ZigBee standard, ZigBee Home Automation devices will interoperate effortlessly with other products already in consumers’ homes using other ZigBee application profiles, such as ZigBee Light Link, ZigBee Remote Control, ZigBee Smart Energy or ZigBee Building Automation.

Finally – the standard is interoperable – integrating control and monitoring devices for lighting, security, home access and home appliances, allowing the customer to select from a variety of different products to meet her needs. All ZigBee-certified products are interoperable with each other and with other ZigBee networks, regardless of their manufacturer. All certified ZigBee devices, including but not limited to ZigBee Home Automation devices, from different vendors all use the same standards and are tested and certified to be fully interoperable with each other, allowing the consumer to purchase new devices with confidence.

With our existing experience in producing a wide range of devices incorporating Zigbee-based wireless technology our engineers can take your ideas for home automation to the final product stage.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

Next in our series examining emerging low-power wireless standards, we consider 6LoWPAN, which stands for “IPv6 over Low-Power Wireless Personal Area Network”. This is a set of networking standards and specifications which is designed to address the ideas that the Internet Protocol (IPv6 in particular) can be and should be applied to even the smallest embedded wireless Internet-of-Things connected devices right out to the “end branches” of the network; and that power-efficient embedded devices with limited processing power should be fully able to be a part of the Internet of Things, including the use of IPv6 network connectivity.

Whilst the Internet Protocol is the workhorse for the Internet and local-area networks, the IEEE 802.15.4 standard defines the networking of wireless mesh devices. Although the two different protocols are inherently different, the 6LoWPAN specification defines encapsulation and header compression mechanisms that allow IPv6 packets to be sent and received over IEEE 802.15.4 wireless networks, essentially allowing the two standards to operate together, efficiently bringing the Internet to small, power-efficient, cheap devices without the relatively high cost, complexity and power consumption required to implement IEEE 802.11 wireless LAN connectivity at every wireless network node.

For example, a typical embedded 802.11 Wi-Fi module may consume 250 mA while it is awake and actively transmitting, and it may well require a separate microcontroller to interface it to the sensors or other electronics required for a particular application. On the other hand, a system-on-chip incorporating a microcontroller combined with an 802.15.4/6LoWPAN-compatible radio transceiver may only consume 25 mA when it is awake and actively transmitting RF data – an order of magnitude less power consumption.

6LoWPAN is well suited to small, compact, relatively low-cost embedded Internet-of-Things appliances that require wireless connectivity to the LAN and to the Internet but can accept connectivity at a relatively low data rate. Examples may include embedded automation, building control systems and wireless sensor networks in home, office and industrial environments, as well as smart energy metering, measurement and control networks. Devices such as smart meters may collate their data via a 802.15.4/6LoWPAN mesh network before sending the data back to the billing system over the IPv6 backbone.

Whilst IP networks are typically designed to optimise speed whilst managing traffic issues such as network congestion, 802.15.4 systems are designed to give a higher priority to efficient low-power operation and optimisation of memory use, maximising their utility on small, cheap, memory-constrained microcontrollers.

There are some complexities involved in interfacing the two systems elegantly – for example, whilst IPv6 requires a maximum transmission unit of at least 1280 bytes, the 802.15.4 physical layer allows a maximum of 127 bytes per packet, including the payload. The management of addresses for devices that communicate across both the dissimilar domains of IPv6 and IEEE 802.15.4 is also cumbersome, as is the routing of packets between the IPv6 domain and the PAN domain.

Since IP-enabled devices may require the formation of ad-hoc networks particularly during initial setup and configuration, the current state of neighbouring devices and the services hosted by such devices will need to be known. This requires a mechanism for device discovery of the neighbouring devices present in the network.

All 802.15.4 networks connected to the Internet, using 6LoWPAN or otherwise, do require the hardware and software of a physical “bridge” or “gateway” at some point or points in the network, in order to connect the 802.15.4 wireless mesh network to an 802.11 wireless LAN or wired Ethernet. Multiple such nodes mitigate the possibility of single-point failure of network connectivity for the mesh network, at the price of increased network complexity and hardware cost.

IPv6 nodes are assigned 128-bit IP addresses in a hierarchical manner, through an arbitrary length network prefix. IEEE 802.15.4 devices may use either 64-bit extended addresses or 16-bit addresses that are unique within a PAN (a Personal Area Network, which is a group of physically colocated 802.15.4 nodes) as long as an association between a node and a particular PAN has occurred. A particular PAN can also be identified by giving it a PAN ID, allowing the devices of that PAN to easily be recognised – for example, a particular PAN may be associated with a particular building or a particular room.

IEEE 802.15.4 is specifically intended for compact, cheap devices with a relatively low power consumption, operating efficiently from power sources such as batteries. After all, for networks of numerous Internet-of-Things appliances to become ubiquitous, individual wireless hardware nodes need to be as compact, unobtrusive and as cheap as possible.

Making each hardware device as small as possible also allows for portability and greater flexibility in how the devices are used – in wearable computing, for example. However, devices that don’t need to be wireless can be kept in the IP domain of the network and wired in to copper Ethernet – and if portability isn’t required, this means more bandwidth is available to the device. In such a case, a wired mains power supply may also be used, meaning that a larger amount of power is available.

In applications where wireless networking is required but device cost and power efficiency does not need to be so tightly constrained, or where more network bandwidth is required, 802.11 wireless networking may be chosen instead of 6LoWPAN over 802.15.4, keeping the device in the IP domain.

As you can imagine the 6LoWPAN standard offers new levels of compatibility with upcoming infrastructure and is perfect for low-power applications. And if this meets your needs but you’re not sure how to progress with a reliable implementation, we can partner with you to take care of this either in revisions of existing products or as part of new designs. With our experience in retail and commercial products we have the ability to target your product’s design to the required end-user market and all the steps required to make it happen.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

The Bluetooth wireless data protocol has been in use for over ten years, and in recent time the new low energy standard has been introduced. This gives designers another option for wireless connectivity between devices with an extremely low power consumption. In the following we examine what it is, the benefits and implementation examples.

Bluetooth LE (for “low energy”) is aimed at novel applications of short-range wireless communication in connected Internet-of-Things devices for medical, fitness, sports, security and home entertainment applications, and was merged into the main Bluetooth specification as part of the Bluetooth Core Specification v4.0 in 2010.

Also known as “Bluetooth Smart”, it enables new applications of Bluetooth networking in small, power-efficient Internet-of-Things devices that can operate for months or even years on tiny coin cell batteries or other small-scale energy sources. Bluetooth LE devices offer ultra-low power consumption, particularly in idle or sleep modes, multi-vendor interoperability and low cost, whilst maintaining radio link range that is sufficiently long enough for the intended applications.

The Bluetooth LE protocol is not backwards-compatible with the “classic” Bluetooth – however, the Bluetooth 4.0 specification does allow for dual-mode Bluetooth implementations – where the device can communicate using both classic Bluetooth and Bluetooth LE. Whilst Bluetooth Low Energy uses a simpler modulation system than classic Bluetooth, it employs the same 2.4 GHz ISM band, allowing dual-mode devices to share a common antenna and RF electronics for both Classic and Bluetooth LE communication.

Small, power-efficient devices like wearable athletic and medical sensors are typically based on a single-mode Bluetooth LE system in order to minimise power consumption, size and cost. In devices like notebooks and smart phones, though, dual-mode Bluetooth is typically implemented, allowing communication with both Bluetooth LE and classic Bluetooth devices. When operated in Bluetooth LE mode, the Bluetooth LE stack is used whilst the RF hardware and antenna is usually the same set of hardware as used for classic Bluetooth operation.

Devices using Bluetooth LE typically have a power consumption, for Bluetooth communication, which is a fraction of that of classic Bluetooth devices. In many cases, devices can operate for a year or more on a single coin cell. This potentially makes Bluetooth LE very attractive for Internet-of-Things networks, telemetry and data logging from environmental sensor networks, for example.

Since many modern consumer devices such as mobile phones and notebooks have built-in Bluetooth LE support, data can be delivered directly to the user’s fingertips from the Bluetooth sensor network with no need for an intermediary gateway or router as would be required for an Internet-of-Things network employing other technologies such as 802.15.4 ZigBee. This direct interoperability with a large installed base of smart phones, tablets and notebooks could potentially be a very significant attraction of Bluetooth LE networks in wireless sensor network and Internet-of-Things applications.

An active Bluetooth radio has a peak current consumption on the order of about 10 milliamps, reduced to about 10 nanoamps (ideally) in sleep mode. In a Bluetooth LE system, the objective is to operate the radio with a very low duty cycle on the order of about 0.1-0.5%, resulting in average current consumption on the order of 10 microamps. At an average current consumption of 20 microamps, such a system could be operated off a typical CR2032 lithium coin cell (with a charge capacity of 230 milliamp-hours) for 1.3 years without battery replacement.

The lower power consumption of Bluetooth LE is not achieved by the nature of the radio transceiver itself (since the same RF hardware is typically used, in dual-mode Bluetooth devices), but by the design of the Bluetooth LE stack to allow low duty cycles for the radio and optimisation for transmission in small bursts – a Bluetooth LE device used for continuous data transfer would not have a lower power consumption than a classic Bluetooth device transmitting the same amount of data.

The Bluetooth specifications define many different profiles for Bluetooth LE devices – specifications for how a device works in particular families of applications. Manufacturers are expected to implement the appropriate profiles for their device in order to ensure compatibility between different devices from different vendors. A particular device may implement more than one profile – for example one device may contain both a heart rate monitor and a temperature sensor. Here is a non-exhaustive list of a few different Bluetooth LE profiles in use:

Health Thermometer Profile, for medical temperature measurement devices.
Glucose Monitor Profile, for medical blood glucose measurement and logging.
Proximity Profile, which allows one device to detect whether another device is within proximity, using RF signal strength to provide a rough range estimate. This is intended for security applications as an “electronic leash”, allowing the detection of devices being moved outside a controlled area.
Running Speed and Cadence profile, for monitoring and logging athletic performance.
Heart Rate Profile, for heart-rate measurement in medical and athletic applications.
Phone Alert Status Profile, which allows a client device to receive notifications (such as an incoming call or email message) from a smart phone. As an example, this is employed in the Pebble smart watch.

The Bluetooth LE shows a lot of promise, and with a minimal chip set cost gives the designer another cost-effective wireless protocol. And if this meets your needs but you’re not sure how to progress with a reliable implementation, we can partner with you to take care of this either in revisions of existing products or as part of new designs. With our experience in retail and commercial products we have the ability to target your product’s design to the required end-user market and all the steps required to make it happen.

Making the decision to create a new product or the next generation of an existing product is always an exciting time for design engineers and hopefully the entire organisation. There’s always new features, options and technologies that can be integrated for the perceived benefit of the end user.

However as technology marches on, there is the possibility of going too far. At first that may seem like an odd statement, however considering the complexity of some products you may wonder how they’re comprehended by the end-user, let alone sales staff. This phenomena is also prevalent in the Internet-of-things arena, where “features” and usability can get out of hand.

Let’s consider the potential dangers of over-engineering and feature overcomplexity when bringing an Internet-of-Things automation or embedded sensor appliance to the market. With advancements in available technology, increasing miniaturisation and decreasing costs of sensors and components it’s tempting for more and more features and capabilities to be added to your device or product design, just to make your product “the best” or to satisfy the “because we can” motivation of the engineering team.

However, it can be important to keep this kind of over-engineering or “feature creep” under control in order to deliver a product that is easy for consumers and salespeople to understand and offers simple, sensible, intuitive user experience with a sensible amount of functionality – not too little and not too much – from a hardware system that is small enough and simple enough that it can practically be manufactured and offered to the market at an acceptable price for good consumer uptake.

Sure, your design might be “the best” from a technology standpoint, but what if the “best” hardware is significantly more expensive than the competitor’s not quite as whiz-bang product and your design is not considered financially attractive to consumers relative to the level of functionality that the users actually want?

It’s pointless to try and invent more and more features just because it is technologically possible to do so if those features don’t actually accomplish anything that is actually valuable to consumers. For example, providing a washing machine with Internet-of-Things connectivity and remote access and control via email or a smartphone application is quite pointless since a human operator actually needs to be there to load and unload the clothes from the machine.

The features and user experience should be kept intuitive and usable, without dragging the user down into an insane range of different options that most people are probably never going to use most of the time anyway.

Internet-of-Things sensor networks and appliances targeted at home and building automation should be easy to set up and configure, they should be compatible with existing typical household network infrastructure such as single-band 2.4 GHz 802.11b/g Wi-Fi access points (5 GHz might be technically “better”, for example, but users don’t want to upgrade all their existing access points just to use your gadget), and they need to be compact, visually unobtrusive – and as simple as possible in order to keep the hardware cost at a level that is sufficiently small for market acceptance.

This is particularly true for appliances that are designed for use as a network of many distributed devices – the cost of the total set of all the hardware devices needed for a typical network deployment needs to be kept at a reasonable level so that the entire usable system is available to consumers at an overall price point that they’re willing to pay. For Internet-of-Things networks consisting of meshes of multiple wireless devices to become ubiquitous, each node device needs to be as cheap and as small as possible.

For example, suppose that you release a smart email-controlled Internet-of-Things light bulb onto the market and it costs $100. Will customers replace their existing light bulb, which costs say $5, with your new $100 light bulb with the added convenience of control from your PC? Well, some consumers might try a single light bulb or two just to experience the relatively novel idea of a consumer-focussed household Internet-of-Things appliance.

However very few customers are likely to consider it worthwhile to set up a network of a dozen hundred-dollar bulbs to replace every bulb in the home. Such a system might pick up a few customers – the relatively wealthy technology fans who want to be early adopters of advanced, relatively complicated home automation and Internet-of-Things technologies, even if the price is high. But isn’t it better to have a product that is desirable for a broader market beyond just those who are willing to pay lots of money for the most powerful, advanced technology on the market?

Furthermore, realistic testing of your product’s usability and user experience is vital during the development process. Adding too many features can befuddle customers as well as befuddling salespeople whose job it is to help convince customers to buy your product and to demonstrate its user experience with consumers. Over-engineering and feature creep, even if it’s possible to integrate lots and lots of features from a technical engineering standpoint, can negatively affect sales as well as affecting your brand reputation.

The best user interface is “no user interface” – a user interface design that approaches the theoretical ideal of being completely transparent and natural in its interaction with the user. Similarly – although I know it might sound risky – the best documentation design is “no documentation”, or something approaching it. The ideal product is so intuitive and natural in its user experience that it just kind of “documents itself”, with little or no documentation really required. This means that the amount of documentation that the customer needs to read is minimised as well as minimising the amount and the cost of documentation that the manufacturer needs to print for every unit shipped.

With hindsight you can examine your own existing products and that of your competitor’s, and with a fresh perspective perhaps consider how things can be simpler for the end user without sacrificing usability. This is a simple step to initiate, however it can require a total redesign or approach from a fresh set of minds.

As part of our complete product design service, here at the LX Group we can partner with you to work on revisions of existing products or bring new ideas to life. With out experience in retail and commercial products we have the experience to target your product’s design to the required end-user market and all the steps required to make it happen.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

After the initial excitement of generating an idea for a new Internet of Things device, there’s still countless design considerations to take into account – some of which you may not have even heard of. And a fair amount of these will be generated by the needs of specific markets around the world. So let’s consider some of the challenges involved in designing an Internet-of-Things device or appliance and bringing it to the global market.

What are some of the different factors that need to be taken into account when bringing a hardware device to market internationally? The need for multi-voltage off-line power supplies and multi-lingual product manuals are well-known things we’re used to with all our technology products – but with modern Internet-of-Things gadgets employing Internet connectivity, cloud computing and wireless radio-frequency mesh networks, there are some increasingly important factors to consider which may not be as familiar to the design team.

For mains-powered systems, international differences in mains voltage and frequency are an obvious factor to start with to ensure compatibility with the worldwide market. Modern switch-mode power supplies can easily be designed to span the possible worldwide voltage range between 100 V AC and 240 V AC without manual switching or configuration, at grid frequencies between 50 and 60 Hz. However, it should be remembered that the mains voltage is only assured within a tolerance of around plus or minus 10 percent, so an example of a good input voltage specification for a well-designed modern SMPS might be 85-265 V RMS AC at a grid frequency of 50-60 Hz. Extra attention is needed in systems where a clock or timebase is derived from the frequency of the AC grid – in systems of this sort, manual specification of the frequency may be required even if the power supply itself does not care about the AC frequency.

When designing and deploying wireless sensor networks, Internet-of-Things networks and similar modern technologies where radio communication is used, attention also needs to be paid to differing international allocations of RF spectrum and licensing requirements for the use of the RF spectrum. Spectrum allocations and licensing requirements for Industrial, Scientific and Medical (ISM) bands differ between countries – for example, the 915 MHz band should not be used in countries outside ITU Region 2 except those countries that specifically allow it, such as Australia and Israel.

A device that operates with a certain frequency spectrum and power level that requires no license, or falls into a class license, in one country may not be able to be legally operated in another country without specific operator licensing. For example, some devices operating in the 70 cm (433 MHz) spectrum that fall within the Low Interference Potential Device (LIPD) class license in Australia and hence can be freely operated cannot be used in the United States except by licensed amateur radio operators. The European Union’s Reduction of Hazardous Substances (ROHS) directive took effect in 2006, restricting the use of certain substances considered harmful to health and the environment, such as lead and cadmium, except in technological applications where elimination of these elements is not viable.

While RoHS compliance is not required for all electronic equipment sold throughout the world and is only strictly required for devices sold into the EU market, it is achieving widespread acceptance throughout the electronic manufacturing industry worldwide. However, in some specialised applications where extremely high reliability and resilience against factors such as tin-whisker formation is required, such as space and defence technology, these factors may take precedence over ROHS compliance and the use of lead-containing solder alloys and platings may be specified.

Different testing organisations are responsible for setting and enforcing the standards for electrical safety and RF spectrum usage in different countries, and it can be challenging to keep track of the different testing requirements needed before bringing your product to market in every market country.

For example, Underwriters Laboratories is well known in the United States for their role in drafting safety standards and providing compliance testing procedures for safety-related factors, whilst approval from the FCC is required to recognise compliance with RF spectrum and electromagnetic interference requirements – a completely separate thing to safety certification. And for another example, the TUV provides a similar role in the verification of safety-related standard compliance in the German market.

Other social and socio-economic factors that might not be as obvious can affect the user experience your product provides in different customer markets – for example, a device that constantly needs to “phone home” to an Internet-connected service may not function effectively in a country without widely available, or reliable, Internet access. In a situation like this, it may be beneficial to have a system designed to store and buffer its collected data locally on a storage device and only synchronise with an Internet service occasionally when connectivity may be available.

In conclusion, there’s a myriad of not only standards but also operational considerations to take in account when designing your next product for the global market. However don’t let that put you off – the greater the challenge, the greater the possible success. But if you’re not sure about testing, standards, compliance, markets abroad or any other factor – parter with an organisation that does: the LX Group.

Here at the LX Group we have the experience and team to make things happen. With our experience with connected devices, embedded and wireless hardware/software design, and ability to transfer ideas from the whiteboard to the white box – we can partner with you for your success.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

The Internet of Things holds great potential, and much has been written about the final applications and the possibilities – however one major factor of any device is how it will be powered. It’s all very well to have the latest sensors or interactive devices, if they don’t have a suitable power supply. It’s easy to consider a battery – however there’s many more options that can increase lifespan, reduce maintenance calls and therefore the running costs. Let’s review a few options that can provide a portable power supply for your IoT nodes where mains power or cabled-in power supplies are not available.

First, consider solar photovoltaic – a common choice for sensing, control or measurement devices that are located outdoors where sunlight is available, and that consume a relatively small amount of power. For a small, low-power embedded device that receives a reasonable amount of sun each day, a moderately small solar panel is perfectly capable of supplying sufficient power – on average – to run a small, basic wireless network node consisting of a microcontroller, some sensors and an embedded low-power Wi-Fi, Bluetooth or 802.15.4/ZigBee radio transceiver. However this is assuming that the overall system is designed for a reasonable degree of power efficiency.

However, solar power is intrinsically intermittent and is only available on average for a fraction of the day. To allow the system to have access to the current it needs to function when needed, solar-powered wireless devices almost always need to incorporate a small amount of energy storage in the form of a battery or supercapacitor in conjunction with the solar cell. Furthermore, solar cells or solar panels typically have a relatively low output voltage if a small number of cells are used, and their non-linear V-I curve makes it desirable to employ Maximum Power Point Tracking (MPPT) where practical.

This is necessary to keep the system operating near the maximum power point so that the limited energy available is harvested most efficiently. A solar power supply for a remote wireless system ideally tracks the maximum power point of the cell along the V-I curve and is able to charge a small battery or supercapacitor to fill in the demand when sufficient sunlight is not available.

As an example of a controller IC one may use for the power supply in a small solar powered system, the Linear Technology LTC3105 is a high efficiency step-up DC/DC converter that can operate from input voltages as low as 225 millivolts, with a built-in maximum power point controller (MPPC). As well as solar cells, this device is well suited to other low voltage, high impedance energy harvesting transducers such as thermoelectric generators and fuel cells.

Whilst it is not a true maximum power point tracker, the user-programmable maximum power point setting helps to optimise the efficiency of energy extraction from any energy source, such as a thermoelectric pile or a solar cell, where the voltage across the transducer may vary with changing environmental conditions as well as with the load current. The LTC3105 is capable of supplying 70 mA of output current at 3.3V from an input voltage of 1 volt – this is sufficient power to run a small, well designed basic sensor node consisting of a microcontroller, RF transceiver and a sensor or two.

Another type of power supply is known as energy harvesting, made possible by parts such as the Linear LTC3108, which is designed to accommodate energy harvesting from transducers with extremely low output voltages, as low as 20 millivolts. This makes it particularly well suited for use with thermopiles and thermoelectric generators which can generate a very low potential difference from a realistic temperature difference – a potentially convenient energy source for remote sensing in industrial automation or process monitoring in high-temperature systems where wired communications and power are not convenient.

Energy can also be derived from vibration, and using a part such as the Linear LTC3588 – a piezoelectric energy harvesting power supply controller which connects to a piezoelectric crystal to harvest mechanical energy in the form of vibrations from the ambient environment. This IC incorporates a low-loss, full-wave bridge rectifier and is capable of accommodating the rapidly changing AC voltage output and high source impedance of a piezoelectric crystal subject to mechanical stress and converting this energy into a DC current with relatively high efficiency.

Output voltage selections between 1.8V and 3.6V are available with a continuous output current capability of up to 100 milliamps, compatible with a range of modern power-efficient microcontrollers and RF mesh systems-on-chip.

An electromechanical energy harvester of this sort can be employed to provide a continuous source of a small amount of “free” energy for a small, efficient wireless network mote, particularly in applications such as vehicles and industrial machinery where plenty of vibrational energy is available to be harvested in the environment.

Finally, for some systems it is also practical to use just batteries – for example, lithium-ion, lithium-polymer or nickel-metal hydride batteries – and rely on user intervention to simply
recharge and replace the batteries where needed. The batteries may be left internally, inside the device, with the system being plugged into a power supply via a charging port
– perhaps using a low-power standard power-supplying interface such as USB – when the device requires a recharge, as opposed to the traditional method of removing and swapping the batteries.

In this sort of application, battery management and charging ICs such as the Microchip MCP73833 Li-polymer / Li-ion charge management controller can be of use to control the recharge of a Li-ion cell, as can buck/boost converters such as the Texas Instruments TPS63031. A buck/boost converter like this allows a regulated output voltage to be generated from input voltages both higher and lower than the desired output voltage – an output of 500mA at 3.3V, in this case, from an input voltage anywhere from 2.4 to 5.5 volts. This allows a battery such as a two-cell NiMH, three-cell NiMH, or single-cell Li-ion / Li-polymer to be used efficiently and charged and discharged across the entire usable part of its discharge curve without hitting the minimum input voltage of a LDO or buck regulator.

No matter what your device or where it will be located, finding an appropriate source of power is possible, and easier than you realise. It just takes a little research and a team of dedicated engineers with the experience and knowledge to understand your requirements. Here at the LX Group we have the experience and team to make things happen. With our experience with connected devices, embedded and wireless hardware/software design, and ability to transfer ideas from the whiteboard to the white box – we can partner with you for your success.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

The fact that the Internet of Things shows a lot of promise both now and in the future is certain – and your customers, designers and the public will have an almost limitless amount of ideas with regards to new products and their implementation. However when the time comes to select the hardware to drive these innovations, choosing from one of the wireless chipsets can be a minefield – and more so when WiFi is involved.

802.11 wireless LAN is an attractive technology for building networks of wireless sensors and embedded devices due to its widespread use and the availability of nearly ubiquitous existing network infrastructure. Let’s take a look at a few existing chipsets on the market today that can be used to add wireless networking to existing embedded designs with relatively low complexity and cost.

First there’s the RN131 802.11b/g WiFi module by Roving Networks – a complete low-power embedded networking solution. It incorporates a 2.4 GHz radio, processor, TCP/IP stack, real-time clock, crypto accelerator, power management and analogue sensor interfaces into a single, relatively power-efficient module. In the most simple configuration, the hardware requires only 3.3V power, ground, and a pair of serial UART lines for connection to an existing microcontroller, allowing wireless networking to easily be added to an existing embedded system.

The module incorporates a U.FL connector for connection of an external antenna, without any microwave layout or design needed to use the module. This module has a current consumption of 40mA when awake and receiving, 200mA when actively transmitting, and 4µA when asleep, and the device can wake up, connect to a WiFi network, send data, and return to sleep mode in less than 100 milliseconds. This makes it possible to achieve a runtime on the order of years from a pair of standard AA batteries – an ideal solution for power-efficient, battery powered wireless sensor network and Internet-of-Things solutions.

Next there’s the Texas Instruments CC3000 Wireless Network Processor – which allows WiFi to be added to any existing microcontroller system relatively easily, and at a low cost. The CC3000 integrates an entire IPv4 TCP/IP stack, WiFi driver and security supplicant on the chip, making it easily portable to lightweight microcontrollers without the memory burden of implementing a TCP/IP stack in the host microcontroller where relatively low-power, low-cost microcontrollers such as 8-bit AVR or PIC devices are used. And this compact module measures only 16.3mm x 13.5mm.

CC3000 reference designs available from TI demonstrate chip-antenna based designs that are already FCC, IC and CE certified, which can make it easier to develop bespoke solutions that can pass compliance testing for products going into markets where such compliance is needed. The CC3000 requires no external crystal or antenna balun, and in fact requires almost no external components at all except for an SPI interface to the host microcontroller and an antenna – and the device costs less than $10.

The flexible 2.7-4.8V power supply requirement offers great flexibility when combined with battery power or energy harvesting solutions. However, this chip is not a PCB-based module, meaning that a 50 ohm 2.4 GHz antenna must be added externally – so the designer must have a little familiarity with microwave design, such as microstrip transmission line layout and the choice of the right antenna connector. However, this offers the designer complete flexibility to choose the most appropriate antenna type for the size, range and gain requirements of the design – a larger external antenna, a compact chip antenna, or a microstrip antenna fabricated on the PCB with no bill-of-materials cost.

Our final subject is the Redpine Signals’ Connect-IO-n series of modules which allow 802.11 wireless LAN connectivity to be added relatively easily to an embedded microcontroller system. In collaboration with Atmel these modules have been optimised for use with Atmel microcontrollers, particularly the Atmel AVR XMEGA and AVR UC3 series microcontrollers.

Some modules in this family provide 802.11a/b/g/n Wi-Fi connectivity, whilst all modules provide the TCP/IP stack on board and are FCC certified, making RF compliance certification of your entire design easier. These modules are aimed at providing the ability to add 802.11 wireless connectivity to 8-bit and 16-bit microcontrollers with low integration effort and low memory footprint required in the host microcontroller to support the WiFi device, especially where 802.11n support is desired.

Like the other chipsets we’ve discussed, the modules in this series can be interfaced to the host microcontroller over a UART or SPI interface, and similarly to their competitors, a standby current consumption of only a few microamps potentially allows for years of battery life with no external energy source as long as the radio is only briefly enabled when it is needed.

The RedPine RS9110-N-11-28 module from the Connect-IO-n family in particular is relatively unusual in that it provides dual-band 2.4GHz/5GHz 802.11 a/b/g/n connectivity for your embedded device – supporting connection to any WiFi device or network and potentially avoiding congestion in the 2.4 GHz band as used with 802.11b/g devices.

Whilst 802.11n offers a significant increase in the maximum net data rate from the 54 MBit/s of 802.11b/g to 600 Mbit/s, do you really need 600 Mbit/s of data to your wireless sensor network or embedded appliance? I doubt it. However, one case where you might want an 802.11n radio supporting operation in the 5 GHz spectrum for your wireless sensor network device is if your wireless LAN infrastructure is a pure 5 GHz 802.11n network – whilst this breaks compatibility with legacy devices, it delivers maximum network performance.

As you can see the possibilities for low-power connected devices are plentiful and the hardware is available on the open market. It’s then up to your team to turn great ideas into great products. Furthermore modifying existing products to become connected is also a possibility. However if wireless or Internet-connectivity is new to your team – and you’re in a hurry, have a reduced R&D budget, need guidance or want to outsource the entire project – it can be done with the right technology partner.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.

Wearable computing – the use of personal computers, displays and sensors worn on one’s person – gives us the potential for advancement in human-computer interaction compared to traditional personal computing – for example the ability to have constant access and interaction with a computer – and the Internet, whilst going about our daily activities.

This could be considered the ultimate in multitasking – the use of your computing device at any time without interrupting your other activities. For example, the ability to read an email or retrieve required information while walking or working on other tasks. Wearable computing potentially offers much greater consistency in human-computer interaction – constant access to the computer, constant connectivity, without a computing device being used in an on-and-off fashion in between other activities.

Once contemporary example of this is the new Google Glass, which represents an advanced, sleek, beautifully designed head-mounted wearable computer with a display suitable for augmented-reality applications – or just as an “ordinary” personal head-mounted display. Even before its public release, the frenzy surrounding Google Glass amongst technology enthusiasts demonstrates the potential level of market demand for wearable computers.

However, with a price of at least US$1500 price tag of Google Glass, (at least for its “Explorer Edition” beta version) this leads many to consider what potential might exist for the deployment of wearable computing and wearable sensor-network technologies – however at a lower cost.

One example is the category known as “Smart Watches” such as the Sony SmartWatch and Pebble Technology’s “Pebble” e-Paper watch – which both offer constant, on-the-go access to information from the Internet – and thus become a member of the Internet of Things – at a glance of the wrist. Text messages and email notifications are amongst the most simple, common examples of data that can be pushed to a smart watch, but the display of information from a multitude of other Internet-connected data streams is possible.

With the growing popularity and increasing hardware capabilities of smart phones, it is increasingly taken for granted that a smart phone carried on one’s person can act as a gateway between the Internet (connected via the cellular networks) and other smaller, lower-power wearable computer or sensor devices worn on the body and connected back to the smartphone via standard data links such as WiFi or Bluetooth. In using the smart phone as an Internet connection, the size, price and weight of the wearable device can be significantly reduced – which also leads to a considerable reduction in cost.

Furthermore, apart from providing mobile Internet connectivity, the smart phone can also provide a large display and an amount of storage capacity – which can be harnessed for the logging, visualisation and display of data collected from a network-connected sensor node wearable on one’s body, or a whole network of such sensor nodes distributed around different personal electronic devices carried on the person and different types of physical sensors around the body.

The increasing penetration of smart phones in the market and the increasing availability and decreasing cost of wireless radio-networked microcontroller system-on-chips, MEMS
sensors and energy efficient short-range wireless connectivity technologies such as Bluetooth 4.0 are among some of the factors responsible for increasing the capabilities of,
and decreasing the cost of, wearable computing and wearable Internet-of-Things and sensor platforms.

Speed and position loggers, GPS data loggers and smart pedometers intended for logging and monitoring athletic performance, such as the Internet-connected, GPS-enabled,
Nike+ system; along with biomedical instrumentation and sensor devices such as Polar’s Bluetooth-connected heart rate sensors are other prominent examples of wearable Internet-of-Things devices which are attracting increasing consumer interest on the market today.

Combined with display devices such as smart watches, smart phones and head-mounted displays such as Google Glass. these kinds of wearable sensors create a complete wearable machine-to-machine Internet-of-Things network that can be self-contained on one’s person. Which leads us to the next level of possibilities – what do your customers want a device to do? And how can it be accomplished? And do you have the resources or expertise to design, test and bring such a system to the market?

It isn’t easy – there’s a lot of technology to work with – however it can be done with the right technology parter. Here at the LX Group we have the experience and team to make things happen. With our experience with sensors, embedded and wireless hardware/software design, and ability to transfer ideas from the whiteboard to the white box – we can partner with you for your success.

Published by LX Pty Ltd for itself and the LX Group of companies, including LX Design House, LX Solutions and LX Consulting, LX Innovations.