List: From AI to machine learning, sensors and nodes, the first results of the government’s £40m IoT boost are here.
In 2015 the British Government announced a £40 million scheme to further the research into IoT connected devices as well as adoption.
This week, IoTUK hosted a Showcase event, in which different start-ups exhibited their smart IoT products aimed at the health and safety space.
CBR lists five products featuring at the event.
Aiseedo: teaching robots like kids
Essentially aiming to be the brain of future robots, Aiseedo has developed an artificial intelligence machine learning system that learns by itself in real-time.
Co-founder Laure Andrieux, said the product is a cloud based SaaS which is easy to use – even for those with no machine learning expertise.
When it comes to improving living conditions, Andrieux said that the system can learn about someone’s living habits, such as an elderly or disabled citizen, and adjust different factors, from temperature to lights, customising the space to the needs of the user.
She said: "With AI you are teaching a robot a movement; you are teaching it like you would teach a kid."
The Aiseedo engine has been designed for streaming real-time data and data fusion, and consists of neural network components that analyse input messages and are architected to accommodate varied, asynchronous data sources.
Andrieux told CBR that the technology has many different applications, not only in the connected living space, but also in the oil and gas space, and other data intensive industries. In the future, the technology could also be used in remote search operations.
"You could tell the system [in the robot] to go down a tunnel and only pick up a certain object," she said.
To showcase this, Andrieux used a live demo with the Cookie Monster using closed loop analytics technology to learn from its sensors about the world around him, navigate it and eat as many cookies as he can find.
The system, which learns in real time, demonstrates how the Cookie Monster understands and reads the obstacles in front of him, such as walls, and either navigates around them, or slows down to avoid crashing into them.
The engine uses a combination of several machine learning techniques including feed forward neural nets, recurrent neural nets, reinforcement learning and data fusion, to extract patterns from the message streams and build a holistic predictive model of the cookie monster’s world.
"In the industrial robotics sector, imagine an object out of place on a production chain: a robot, using artificial intelligence machine learning capabilities could detect, go over and re-arrange that object position before performing its task, just by having naturally learnt that that was the wrong place for the object for the task to go on well" Andrieux said.
Pacif-i: a smart baby dummy
Pacif-i is a smart baby dummy that monitor’s babies’ temperature, medication and the device’s location, all via a mobile app.
The company claims Pacif-i to be the world’s first Bluetooth Smart thermometer pacifier and works on the iOS 7.0 and Android 4.3 and later.
The app can monitor more than one baby’s dummy at a time and is suitable for those between zero and three years old.
The water-resistant product has been designed by British health and environment solutions company Blue Maestro.
In answer to one of the biggest challenges with IoT devices, founder Kristin Hancock said that Pacif-i’s "battery life lasts for the product lifetime, which is around a year," depending on the use. Normal dummies are usually replaced every four to six months.
Retailing at £39.99, Hancock said that she expects the device to be retailing for a lower price in the future.
As for the collected data, Hancock said: "The harvested data is stored in between the owner and the pacifier, not in the cloud."
The device has a reach of 30 ft (pacifier to mobile phone). Medication can be tracked and Pacif-i will let parents know how the baby is reacting to the drug. They can also set up reminders.
BeanIoT: small smart ‘giant beans’
Described as a configurable wearable and deployable IoT edge device, BeanIoT is a small sensor (45x18mm) that can be placed nearly everywhere to monitor room gases, temperature, humidity and so on, as room sensors.
According to Andrew Holland, CEO of RFMOD, the company behind BeanIoT, the devices, which are still waiting on a patent approval, will only report back once measurements go outside limits.
"The sensors built in it are temperature and humidity. It also has an accelerometer, gyro, and a magnetometer," he told CBR.
Holland said that the beans can be used in assisting with independent living. "For example, if you were a care taker, you could put these throughout a small flat. They are the fixed nodes, and they can be used to measure the environment, etc.
"You can then also place one node in the lapel of a coat, and it could detect an angle and or a shock, and so on. This could pick up that someone that has fallen over has not come back up again for a while, it also could monitor a person walking from room to room and understand that it is unusual for that person to be in the kitchen for an hour, for example."
The devices work on a mesh system in the cloud, and encryption is used to secure the product. "We are looking to engage with security experts to ensure that this as an offer is super secure from the outside."
"All the data from the Bluetooth nodes are set up using our app, a third party app, and or system, and you can profile it for your particular application. The data is either transported from Bluetooth to WiFi from a phone to the cloud, but the phone needs to lead the system an any point."
If the router goes down, the beans will continue to collect information and update the data streams once connection is re-established.
Intellicone: smart road works safety
Highway Resource Solutions (HRS) has developed an electronic perimeter management system to monitor roadside worksites and ultimately prevent worker deaths.
According to Roger Poeth, director at HRS, "being a road worker is one of the most dangerous jobs you can have". To put an end to that, the company has built Intellicone, a "smart device that controls temporary work sites" and improves the health and safety aspects of the job.
The product consists of portable site alarms, contact based and contactless sensors, which enable the deployment of an electronic perimeter around a worksite with up to a reach of 100 meters, according to Poeth.
Contact sensor technology for motion and impact has been integrated into static and sequential cone lamps in partnership with Unipart Dorman. The company has made available contactless beam sensors from a range of 3.5 meters up to 50 meters.
Poeth said: "[Intallicone] keeps the data waves very low; in a month, for example, you might just pick up some mbs of data.
"The device picks up on 2D, 3G, wifi, etc. It has to be [doing its job connected to the] internet. We tried RFID but with different characteristics of the terrain, the connection was not great."
The product has rechargeable batteries inside and in case of emergency sends out audible alerts. In 2015 to date, is has been deployed on over 1,000 sites with close to 15,000 unique equipment deployments.
Poeth also said that the company is currently working on a geofencing system
Path Feel: smart assisted walking
Healthcare company Walk With Path has developed a wearable technology to prevent people with disabilities and other vulnerabilities, including multiple sclerosis, Parkinson’s disease and diabetes, from falling over.
The company’s insole, Path Feel, provides haptic feedback using vibration to people as they walk. This means that the insole increases people’s natural sensation of the floor they are walking on.
For example, this will help people with neuropathy, which causes people to experience numbness or weakness as peripheral nerves lose their original capacities to feel the flooring, to walk more independently.
Path Feel has been designed with sensors embedded into the wearable, which then measure the pressure under the sole of foot in three different locations (fore, middle and rear foot).
The vibration intensity will depend on the amount of pressure detected by the sensors.