The term artificial intelligence (AI) has been in exis­tence for over 60 years. During that time, research has been conducted into systems and methodologies capable of simulating the mechanisms of intelligent human behaviour. It might sound quite simple, but it has to date posed major challenges to scientists. Because many tasks that most people would not even associate with “intelligence” have in the past caused computers serious problems: understanding human speech; the ability to identify objects in pictures; or manoeuvring a robotic vehicle around unfamiliar terrain. Recently, however, artificial intelligence has been making giant strides, and is increasingly becoming a driver of economic growth. All major technology companies – all the key players in Silicon Valley – have AI departments. “Advances in artificial intelligence will allow robots to watch, learn and improve their capabilities,” said Kiyonori Inaba, Board Member, Executive Managing Officer and General Manager of Fanuc.

Simulating the human brain

Findings from brain research, in particular, have driven advances in artificial intelligence. Software algorithms and micro-electronics combine to create neuronal networks, just like the human brain. Depending on what information it captures, and how it evaluates it, a quite specific “information architecture” is created: the “memory”. The neuronal network is subject to continuous change as it is expanded or remodelled by new information. The technological foundations for state-of-the-art neuronal networks were laid back in the 1980s, but it is only now that powerful enough computers exist to permit the simulation of networks with many “hidden layers”.

Becoming ever better by learning

“Deep Learning” is the modern-day term used to describe this information architecture. The concept involves software systems which are capable of reprogramming themselves based on experimentation, with the behaviours that most reliably lead to a desired result ultimately emerging as the “winners”. Many well-known applications, such as the Siri and Cortana voice recognition systems, are essentially based on Deep Learning software. “Deep Learning will greatly reduce the time-consuming programming of robot behaviour,” asserts Kiyonori Inaba. His company Fanuc has integrated AI into its “Intelligent Edge Link and Drive” platform for fog computing (also referred to as edge computing). The integrated AI enables connected robots to “teach” each other, so as to perform their tasks more quickly and efficiently: whereas one robot would otherwise take eight hours to acquire the necessary “knowledge”, eight robots take just one hour.

New algorithms for faster learning success

Ever-improving algorithms are continually enhancing the ability of machines to learn. As one example, the Mitsubishi Electric Corporation recently launched a quick-training algorithm for Deep Learning, incorporating so-called inference functions which are required in order to identify, recognise and predict unknown facts based on known facts. The new algorithm is designed to aid the implementation of Deep Learning in vehicles, industrial robots and other machines by dramatically reducing the memory usage and computing time taken up by training. The algorithm shortens training times and cuts computing costs and memory requirements to around a thirtieth of those for conventional AI systems.

Special chips for Deep Learning

To obtain the extremely high computing power required in order to create a Deep Learning system, current solutions mostly involve so-called GPU computing. In this, the computing power of a graphics processor unit (GPU) and the CPU are combined. CPUs are specially designed for serial processing. By contrast, GPUs have thousands of smaller, more efficient processor units for parallel data processing. Consequently, GPU computing enables serial code segments to run on the CPU while parallel segments – such as the training of deep neuronal networks – are processed on the GPU. The results are dramatic improvements in performance. But the development of Deep Learning processors is by no means at an end: the ­“Eyriss” processor developed at the Massachusetts Institute of Technology (MIT), for example, surpasses the performance capability of GPUs by a factor of ten. Whereas large numbers of cores in a GPU share a single large memory bank, Eyriss features a dedicated memory for each core. Each core is capable of communicating with its immediate neighbours. This means data does not always have to be routed through the main memory, so the system works much faster. Vivienne Sze, one of the researchers on the Eyriss project, comments: “Deep Learning is useful for many applications, such as object recognition, speech or facial recognition.”

The post AI – the Learning Robot appeared first on Future Markets Magazine.

The consequences of the trend towards ever more online shopping are seen in many towns and cities, especially in the run-up to Christmas. It often seems that there are parcel service vans standing by the roadside every few metres, usually double-parked too – an irritating obstruction that causes congestion on the roads and increases pollution. So experts are working feverishly, and with great creativity, to come up with new, sustainable solutions.
With a little luck, it might be possible to experience one such solution in operation around the Ottensen area of Hamburg. For a period up to March 2017, logistics service provider Hermes Germany has been testing delivery robots from start-up company Starship Technologies. “Using robots can revolutionise parcel delivery, particularly in urban areas,” asserts Frank Rausch, CEO of Hermes Germany.

Using the pavement

The delivery robot 6D9 is a six-wheeled vehicle, 50 centimetres tall and 70 centimetres long. It incorporates a secure compartment capable of carrying a payload with a total maximum weight of 15 kilograms. The delivery robot runs solely on the pavement, at a maximum speed of 6 km/h – that is to say, at walking pace. It only crosses cycle paths and roads after checking first that all is clear. The built-in cameras and sensors ensure that approaching obstacles are automatically detected, and the robot immediately stops. Bright LEDs mean every robot is clearly visible from a distance.

A new service channel

The robots can operate within a radius of up to five kilometres. This means automated deliveries can be made within 30 minutes of a customer placing an order. During the pilot, the delivery robots run back and forth between the participating parcel shops and selected test customers. The consignments carried are regular orders which customers have requested to be delivered to a Hermes parcel shop rather than to their home. Instead of going to the shop to collect their order in person, the test customers can use their smartphone to arrange for a robot to then bring the parcel to their home. So the robots do not replace conventional parcel delivery routes, or indeed parcel services. Rather, with the Starship robot Hermes is piloting a new service channel which eliminates the need for customers to come and collect their order in person from a shop.

Well protected against theft

The consignments in the transport compartment are protected against unauthorised access by a security lock, a surveillance camera and a PIN code. Once the robot has reached its destination, the recipient receives a text message notification and can go to the door to take delivery of the order. The transport compartment is opened using an encrypted personalised link. If any attempt is made to open the compartment by force, the robot immediately triggers an alarm and notifies the operator. The robot’s position can be tracked at all times by its continuous GPS signal.

Still with human assistance

6D9 navigates by a combination of locating signals (such as GPS) and visualising its surroundings based on multiple cameras. The system automatically recognises pedestrian crossings and traffic lights. This is done by means of sensors and nine camera lenses, which convert the received image data fully automatically into appropriate commands in real time. For guidance at tricky spots, and if any uncertainty arises, a human remote operator located at the Starship control centre in Tallinn can connect to the robot over the Internet. The operator is able to view the camera images from the robot and also receives its navigation data, so as to help it out of a difficult situation by remote control. With every trip the parcel delivery robot “learns” more and more, so continually enhances its autonomy. Nevertheless, on their pilot runs the robots are permanently accompanied by a human guide, in order to gather as much information as possible on how they operate.

(Bildnachweis: Istockphoto: Altayb)

The post Parcel delivered by 6D9 appeared first on Future Markets Magazine.

Since September 2016, logistics service provider Fiege has been employing three Toru Cube mobile robots at its footwear warehouse in Ibbenbüren, Germany, to pick online orders. Using its laser and camera system, the perception-controlled, network-connected robot from Munich-based start-up Magazino can autonomously localise and identify individual objects on shelves, pick them, and then transport them to their intended destination.

Flexible working with no rigid programme

Jens Fiege, Executive of the family business of the same name, comments: “Our aim in deploying new technologies is always to make our logistics processes faster and more efficient.” In fulfilling those aims, the robots the company operates possess a high degree of autonomy. When a robot sets off on a job, it does not yet know in detail what it is going to do, but instead decides on the way to its destination. Only once it is close enough to the storage location to actually see the ordered shoes does it make the final adjustments. It determines the position of the box, and from it derives its subsequent actions. Essentially, the robot does not follow a rigid programme, but rather applies a set of rules governing its behaviour which tell it what action is required under specific given conditions.

Robot detects individual objects

A further special feature of the picking robot is that it is not only able to pick complete load carriers, such as pallets, from the shelf, but even individual product items. Its laser sensors and 3D camera system enable it to autonomously detect individual objects on the shelf. Magazino has developed a proprietary object recognition method for the purpose. The system, known as Sheet-of-Light, projects a cross laser comprising two perpendicular laser lines onto the object it is detecting. A 2D camera captures the reflected laser beams and gauges the object based on the position of the lines in the camera image. The method is designed for cuboid objects. Curved surfaces, such as the spines of books, can also be detected. Fewer 3D points are generated than with a 3D camera, so much less computing power is required, meaning the algorithm can be run on a mini-computer for example.

“Our aim in the deploying new technologies is always to make our logistics processes faster and more efficient”
Jens Fliege, Executive, Fliege Logistik

Humans and robots working in parallel

Frederik Brantner, Co-founder and Commercial Director of Magazino, stresses: “It was important right from the beginning that the robots should be able to work in parallel with people. That means part of the picking process can be automated in a flexible, gradual way.” Its safety laser enables the robot to detect not only obstacles in its way but also human employees around it, while also orientating itself within the warehouse. There is no need for reflectors or marker lines on the warehouse floor. Once taught, the connected robot can also share with new robot “colleagues” self-­created maps of its surroundings, as well as experience in handling specific objects or meeting particular challenges, by way of its wireless connectivity. The smart robot is not only capable of working with existing shelving systems, it can also adapt to new situations and changes within the warehouse.

(Picture Credits: Istockphoto: Martinina)

The post Untiring warehouse operative appeared first on Future Markets Magazine.

It’s no longer just the stuff of fantasy to imagine robots fighting on battlefields around the world. Nowadays, they are already exploring dangerous areas, defusing mines or recovering the wounded like the robotic medic Bear does. The Battlefield Extraction-Assist Robot developed by Vecna Technologies was conceived to rescue injured soldiers from combat zones without risking the lives of human medics. Work is currently in progress to integrate more autonomous capabilities into the robot. “The current generation of robots is dedicated to specific missions, and most are still remotely controlled,” says Thierry Dupoux, Research and Technology (R&T) Director at French manufacturer Safran Electronics & Defense. With its eRider, the company has developed a transport platform for the military that is reminiscent of a large quad bike. The vehicle can be controlled by a soldier – or move autonomously through the terrain. “We are using a complementary approach, inspired by the auto industry, which entails the rational and gradual introduction of autonomy functions. This approach can be applied to any modern transport, intelligence or combat platform.”

Robots still need firing orders

The trend towards increasingly autonomous combat robots is even viewed critically by military forces themselves, however. Up to this point, no army in the world has deployed robots that can make a decision to shoot on their own. This also applies for the drones that are currently the object of public concern. Robots like the SGR-1 – which is manufactured by a Samsung subsidiary, armed with a machine gun and deployed by South Korea to monitor the border with the North – require an explicit order from a human to fire. The weapons on the Maars military robot developed by QinetiQ are also triggered by remote control. Yet in terms of the technology, it is not too much of a stretch to imagine that combat robots will use their weapons entirely independently in the future – the decision as to whether robots should be granted this power is no longer a technical question, but a purely ethical one.

Relevant for the UN Weapons Convention

The annual conference to review the UN Weapons Convention in Geneva at the end of 2016 decided to address the topic of fully autonomous weapons that can select and attack their targets without any significant human intervention. “The governments meeting in Geneva took an important step towards stemming the development of killer robots, but there is no time to lose,” says Steve Goose, Arms Director of Human Rights Watch and Co-founder of the Campaign to Stop Killer Robots. “Once these weapons exist, there will be no stopping them. The time to act on a pre-emptive ban is now.”

(Picture Credits: Vecna Technologies)

The post Robots in the military appeared first on Future Markets Magazine.

The interest in automated solutions has grown massively in the cleaning industry over the past few years. Cleaning robots can help the industry counter increasing cost pressure, personnel shortages and the customary high turnover of staff more effectively. What’s more, cleaning robots enable practically flawless results and improved productivity in light of the increased demand for results-driven cleaning. In 2017, Nilfisk – one of the leading manufacturers of professional cleaning appliances – will address this demand with the launch of the Advance Liberty A 50: a scrubbing robot which autonomously wet-cleans floors.

New prospects for the cleaning industry

The cleaning robot is the first result yielded by the Horizon programme – a joint venture between Danish firm Nilfisk and Carnegie Robotics, an American manufacturer of modern robotics sensors and software. The aim of the Horizon programme is to launch autonomous cleaning solutions onto the market and allow customers to clean floors precisely and reliably without any kind of human operation. “With this programme, we are paving the way for the long-term, strategic development of autonomous, networked cleaning solutions. Thanks to state-of-the-art technology, productivity and overall operating costs need to be viewed from an entirely new perspective,” says Jonas Persson, CEO of Nilfisk.

Military and space-grade technology built in

The cleaning robot incorporates a system comprising sensors, cameras and software, using which it can recognise a room by passing through it just once. Even obstacles the size of a tennis ball can be detected and independently avoided by the robot. This means that it can also be deployed on busy surfaces or during the opening hours of supermarkets, for example. The combination of sensors, cameras and lasers also enables the scrubbing robot to clean close to obstructions and walls. The sensor systems can also operate in low ambient light levels, meaning that the rooms which are to be cleaned don’t even need to be brightly lit. That saves on energy costs. “We’ve adapted military and space-grade technologies to provide the Advance Liberty A 50 scrubber with state-of-the-art perception and intelligent navigation that deliver safe and reliable floor cleaning,” says Steve DiAntonio, CEO of Carnegie Robotics. “At the same time, we’ve engineered a simple-to-use interface that enables flexible and efficient operation.”

Cleaning made simple

In the name of simplicity, the robot features just three buttons for setting the cleaning mode: in a building that’s still unfamiliar, the operator can first switch it to manual mode. While standing on the machine, they can guide the robot once through the entire room. In “Fill-in” mode, it’s enough just to delineate the outline of the surface. The robot subsequently cleans the entire surface and dodges every obstacle; even ones which have appeared in the meantime. Alternatively, the operator can also plot a specific course in “CopyCat” mode to show the robot which route it should follow when cleaning the room. “The Advance Liberty A 50 is our most important product innovation yet,” emphasises Nilfisk CEO Jonas Persson. “It will set the standard and lead the way for intelligent equipment going forward in the commercial cleaning industry.”

The post Smarter cleaning appeared first on Future Markets Magazine.

The use of robotic systems in operations is no longer something new; the da Vinci Surgical System has been used in hospitals around the world for several years already, to name just one example. Until now, however, the machines have still been controlled by doctors, making the systems in use little more than remote-controlled tools. Yet now there is a robot which can perform operations autonomously – that is to say, it does not require any input from doctors. The “Smart Tissue Autonomous Robot” (STAR) developed by researchers at the Children’s National Health System in the USA is even able to operate with greater precision than a human surgeon. While the robot needed more time to fulfil the task than its human counterpart, the results were extremely impressive. Surgeons and scientists from the Sheikh Zayed Institute in Washington demonstrated the robot’s capabilities during operations on dead pig tissue and living animals. STAR planned and implemented the suture autonomously under the watchful eye of a doctor.

Improving the quality of operations

STAR was specifically developed to operate on soft tissue such as muscles, tendons, ligaments, nerves, blood vessels and connective tissue. Around 44.5 million surgical procedures are currently carried out on soft tissue every year in the USA. “Our results demonstrate the potential for autonomous robots to improve the efficacy, consistency, functional outcome and accessibility of surgical techniques,” says Dr Peter C. Kim, Vice President and Associate Surgeon-in-Chief, Sheikh Zayed Institute for Pediatric Surgical Innovation. “The intent of this demonstration is not to replace surgeons, but to expand human capacity and capability through enhanced vision, dexterity and complementary machine intelligence for improved surgical outcomes.”
Previously, operations on soft tissue were entirely manual activities and were not able to be supported by robots. The main reason for this is that operations on soft tissue constantly entail unpredictable, elastic and plastic changes which require continuous adaptations on the part of the surgeons.

The system adapts in real time

STAR solves this problem with the help of a tracking system that combines near-infrared fluorescence markers and 3D light-field cameras. Plenoptic cameras also record the direction of incident light rays alongside the usual two image dimensions. This additional dimension affords plenoptic images extra information about the picture depth. Using this system, it is possible to precisely detect any movements and changes in the tissue in the course of the surgical procedure. The tracking system is combined with a smart algorithm which guides and controls the procedure, all the while performing real-time adaptations to changes in the tissue. What’s more, STAR is equipped with pressure-­sensitive sensors, sub-millimetre positioning and powered surgical tools. Its robotic arm features a lightweight construction that permits movements with eight degrees of freedom.
Now that the robot has proven its efficacy, Dr Kim says that the next step is to reduce the size of the surgical tools further and improve the sensors in order to expand the range of application for the STAR system. He anticipates that the system – or certain aspects of the technology – could be ready for use in hospitals within two years.

The post Ultra-precision surgery appeared first on Future Markets Magazine.

Since his third tour of duty in Afghanistan, Romy Carmago has been paralysed from the shoulders down. But Romy has not given up. Although the doctors told him he would never again be able to breathe without a ventilator, he now doesn’t need one. The doctors have also told him he will never walk or be able to use his arms again, but he is sure he will do it one day.

Getting back on their feet thanks to technology

Using technology if necessary: Romy Carmago has established the “Stay in Step” rehabilitation facility near to his home city of Tampa, Florida. The centre offers programmes and provides technical aids with which paraplegics can stand again and perform exercises. It literally helps people get back on their feet.
In late 2016 Romy had the opportunity to try out a totally new piece of kit: the “Human Support Robot” (HSR) from Toyota. The Japanese group’s research institute has supported Romy Carmago’s rehab centre from the very beginning. So it was no surprise that it chose to trial its new care robot, launched in 2015, as a means of assisting paraplegic patients at the centre.

Help for an aging society

The HSR was originally designed to help care for the elderly. It is a lucrative market, particularly as the World Health Organisation (WHO) forecasts that 22 per cent of the global population will be over 60 by the year 2050. The trend in Japan is even more dramatic: almost 40 per cent of its population is predicted to be 65 or older by 2060. So demand for long-term carers is correspondingly high. And that is also the reason why so many care robots are being developed in Japan.
This is also where the versatile HSR seeks to help. By taking over everyday tasks, it enables people who need care or rely on assistance to continue living an independent life at home. The lightweight (37 kilogram) robot, standing about a metre tall, can use its flexible gripper arm to pick up objects from the floor or from a shelf and put them back again, and open and close curtains for example.
The HSR can also be controlled remotely. Family or friends can operate the robot even when they are away from the location. When doing so, the remote operator’s face is shown on the display and their voice is heard in real time, so enabling the person being assisted to interact with family and friends.
In order to drive the ongoing development of the robot, Toyota has established the HSR Developers’ Community in conjunction with a number of research institutions. Its aim is to put the HSR through practical trials, assure its continuous improvement, and achieve its launch as rapidly as possible. The robot is also being made available to various partner organisations such as universities, and also care homes, in order to refine the software further.

Practical trial

And so it was that the HSR also found its way to Romy Carmago, who tested the robot in his home. The assistance robot was required to perform two quite simple tasks in helping Romy with his day-to-day routine: open the house door as soon as he approached; and give him a drink when needed. They were two tasks that took a lot of training for the little robot. The appraisal by Toyota’s research team at the end of the trial is anything but euphoric: they note that there is still a long way to go before the technology is fit for the real world. But the development work goes on. The researchers are confident that the HSR will one day possess the necessary capabilities to enable people like Romy Camargo to live a more independent life.

The post Help to get people back on their feet appeared first on Future Markets Magazine.

A mission to Mars is one of NASA’s loftiest goals, yet robots are set to do the dangerous groundwork before humans set foot on the Red Planet. “Advances in robotics, including human-robotic collaboration, are critical to developing the capabilities required for our journey to Mars,” says Steve Jurczyk, Associate Administrator for the Space Technology Mission Directorate (STMD) at NASA Headquarters in Washington. With this aim in mind, the US space agency has been working for several years to develop the R5, a humanoid robot constructed by the Johnson Space Center. Its design featuring two legs, two arms and a head allows the robot to work alongside humans or perform high-risk tasks in their place.

Versatile Valkyrie

NASA named its model “Valkyrie” after the messenger to the gods in the Norse sagas. The robot is around 180 centimetres tall, weighs 136 kilograms and walks almost like a human does. 28 torque-monitored joints allow it to perform a wide range of movements. Each upper arm alone is equipped with four elastic rotary actuators arranged in series and has seven joints together with the lower arm. One further rotary actuator allows the robot to rotate its wrist, whilst linear actuators control the tilt and yaw angle. A simplified humanoid hand with three fingers and a thumb ­enables it to grip different objects. This means that the R5 can even turn a door handle. Three additional rotary actuators are accommodated in the robot’s pelvis to control movement in its waist and hip joints.

More than 200 sensors

The robot perceives its surroundings using a multi-modal sensor made by the firm Carnegie Robotics: the system collects distance data using a laser scanner and a stereo camera and supplements these with a video image of its surroundings. Further cameras are built into the torso. More than 200 sensors produce additional information; each hand alone is equipped with 38 sensors (six on the palm and eight along each of its four digits). Valkyrie processes this array of data using two Intel Core i7 COM Express processor cores.

Robots don’t always work

A little more development work is still needed before this robot can actually be sent to Mars, however. NASA has therefore tasked three universities – among others – with improving certain functions. As a consequence, work is being carried out on the autonomous functions, environmental perception and movement optimisation at MIT, the University of Edinburgh and Northeastern University in Boston. Sarah Hensley, an MIT student who is working on the elbow control system, understands only too well why this work is necessary. When Valkyrie is turned on and moves, Hensley says, it often “kind of shivers and falls down. Sometimes robots work, and sometimes they don’t. That’s our challenge.”

A million dollars for the winner

NASA therefore wants to hold a competition in order to provide additional incentives for improving Valkyrie. A million dollars are offered to anyone who succeeds in demonstrating that a virtual model of the R5 can repair the damage done to a Mars habitat by a sandstorm in a digital environment. To be specific, it would need to align a satellite dish, repair a solar energy system and patch up a leak in the habitat. The winners will be announced at the end of June 2017. The software developed in this competition should be transfer­able to other robot systems. In this way, the new technology is intended to benefit older robots and future systems alike. “Precise and dexterous robotics which are able to work with a communications delay could be used in space flight and ground missions to Mars and elsewhere for hazardous and complicated tasks, which will be crucial to support our astronauts,” says Monsi Roman, Program Manager of NASA’s Centennial Challenges.

The post Mission to Mars appeared first on Future Markets Magazine.

Somewhere between Amsterdam’s Schiphol airport and the town of Haarlem, on the longest express bus route in Europe: a bus stands at a traffic light, waiting for the signal to move off. At this light, two horizontally adjacent red dots mean stop; two vertically arranged white dots mean go. The light switches to white, the bus smoothly sets off, and pulls into its lane. Red light ahead – the bus brakes safely and gently, and comes to a stop. In all these manoeuvres the bus is controlled not by a human driver, but by electronics. It is, in fact, the world’s first self-driving city bus to operate in real traffic.

Autonomous driving

There is still a driver on-board, but he is somewhat underemployed as the bus continues on its journey, even as it negotiates two bridges and an underpass. The bus keeps safely in lane. As it reaches the outskirts, it accelerates up to the 70 km/h limit. The maximum speed is pre-programmed, and even at this speed the driver is doing nothing. The bus comes to a halt at the stop, the doors open and close, the bus sets off again – all automatically. At the next traffic light, the bus’s high-tech camera system detects the position of the signal. The bus also communicates with the road infrastructure and receives information on the status of traffic lights via its on-board wireless network. This enables it to run through all the lights as they automatically switch to green for it. As the light changes, there are still pedestrians crossing the street. The bus waits until they have crossed and the road ahead is clear before setting off. The bus has an automatic braking system which activates as and when necessary to avoid collisions.

Developed specially for cities

Daimler has been trialling this so-called Future Bus in real traffic on a BRT (Bus Rapid Transit) route around Amsterdam since 2016. Daimler has been working for a number of years to advance self-driving technology to production maturity. The new Mercedes-Benz E-Class, for example, was the world’s first mass-produced car to be granted a test licence for autonomous driving, in the US state of Nevada. And the Highway Pilot developed by the Daimler Trucks division is currently being trialled as a partially automated driving system for trucks. Dr Wolfgang Bernhard, Member of the Management Board of Daimler AG and Head of Daimler Trucks & Buses, comments: “Launched almost two years ago, our Highway Pilot demonstrates that autonomous driving will make long-distance truck transportation more efficient and safer. We are now bringing the technology to our city buses as the City Pilot. The system has been specially adapted to operate in cities. It runs in partially autonomous mode in specially assigned bus lanes.” BRT networks such as in Amsterdam are ideal candidates as a first step towards fully automated city bus operation: an unchanging route, running in a separated lane; a clearly defined timetable; unambiguous and identical actions at stops.

Cameras and sensors intelligently interlinked

The self-driving bus detects whether the route ahead is suitable for automated running, and signals the fact to the driver. A press of a button by the bus driver activates the City Pilot. For the system to operate, the driver must lift his foot off the accelerator or brake pedal and not steer, because any manual action would override the system. So the driver always retains control, and can intervene as necessary. The autonomous driving solution incorporates the latest assistance systems, as are used on-board Mercedes-Benz coaches for example, as well as additional systems which have been partially adopted by Daimler Trucks and enhanced to handle urban traffic. They include long and short range radar, a multiplicity of cameras, and the GPS satellite navigation system. Cameras and sensors are intelligently interlinked to create a detailed picture of the surroundings and identify the bus’s exact position.

Legal framework still lacking

The partially autonomous city bus aims to significantly enhance safety in urban traffic. Experts predict that autonomous driving will cut the number of accidents by 80 per cent by 2035. The predictive capabilities of the Future Bus also improve efficiency, place less strain on components, and reduce both fuel consumption and emissions. And the smooth, uniform ride additionally enhances passenger comfort. The law does not yet allow regular autonomous driving on the road however. “We urgently need to adapt the rules of the 20th century to the 21st,” Dr Bernhard warns. “But we should not get bogged down in bureaucracy. Before we start debating all the potential issues linked to fully autonomous driving, we first need to make partially autonomous driving possible. We need to allow drivers to take their hands off the wheel.”

The post Hands off the wheel appeared first on Future Markets Magazine.

On the Costa Diadema, the flagship of the Costa Crociere line, guests are now welcomed on-board and provided with assistance by five “Pepper” model robots. The Pepper is the world’s first robot capable of read human emotions and needs, and interacting proactively with people.

“Ahoy” in English, French and Italian

The robots developed by Softbank Robotics operate on the Costa Diadema’s seven-day cruises around the western Mediterranean. Their mission: to entertain the guests on-board, and make the cruise even more unforgettable than it would otherwise be. The robot was first launched in Tokyo in 2014. It incorporates lots of innovative features, and has high-level communication skills. It is able to converse faultlessly in English, French and Italian, and can recognise voices and faces. “By our deployment of emotionally interactive robots on-board our cruise ships, we have once again demonstrated our great innovative capability. For us this is a major step towards the digital future of our cruise brands. I am certain our guests will love Pepper,” comments Michael Thamm, CEO of the Costa group.

Recognising human emotions

Pepper is the first humanoid robot in the world capable of recognising the key human emotions: joy, sadness, ­anger or surprise. It can also smile, furrow its brow, as well as interpreting the tone of a voice. Pepper is also able to interpret the words a person uses, as well as non-verbal signs such as a nod of the head. By combining all this information, the robot can determine whether its human counterpart is in a good or bad mood. The robot’s 2D and 3D cameras enable it to find its way around, and object recognition software allows it to recognise faces and objects. A wide range of other sensors complement the camera systems, including ones to measure distances to obstacles for example. Among these are two ultrasonic and six laser sensors, as well as three obstacle detectors in the little robot’s legs. Pepper also has tactile sensors in its hands which it uses when playing games or interacting socially.

Entertaining and informing

The guests on-board the Costa Diadema encounter the robots in various areas of the ship. Pepper is not only able to answer questions put to it, but can also interact proactively as soon as a passenger approaches. Pepper dances and plays with guests, and poses for selfies with them. The robots are designed not only to make the guests smile, however, but also to provide them with useful information about their journey and the amenities available on-board. The robots know everything about the ship’s restaurants, bars and shops, and can give details of shore excursions and on-board activities. They can also provide information on upcoming destinations and the ship’s route. And Pepper also asks the passengers on the Costa Diadema what they think about the cruise and the on-board ­amenities.

Anyone who visits EBV Elektronik’s stand at optoelectronic trade fairs can experience Pepper in action; giving presentations and joking around with visitors. Sebastian Hülck, Director Segment Lighting at EBV, talks about his experience with Pepper.

How long have you been using Pepper?

Sebastian Hülck: We rolled him out for the first time at the LED Professional Symposium 2016 in Bregenz, Austria. I believe we were the first ones to have used Pepper at a European trade fair.

How difficult is it to adapt Pepper for your own applications?

S. H.: It does take a great deal of effort, that much is true. The firm NoDNA in Ismaning took care of that for us: they are one of the authorised dealers for Softbank Robotics/Aldebaran. It took around three months to get Pepper up to his current standard – and to equip him with the various gimmicks which he has mastered to date.

So Pepper can’t be programmed quickly using an app?

S. H.: No. There are specific development environments for Pepper which you have to get to grips with first, such as the Choreographe Suite. In addition to that, the robot is a very complex piece of kit – with his eight-core Intel Atom platform, 20 motors and array of sensor systems. As an operating system, the machine runs the Linux-based Qi OS.

Does Pepper actually recognise human emotions?

S. H.: If you interact with him on a one-to-one basis, that actually works pretty well. On the other hand, he is overwhelmed when surrounded by lots of people. In that scenario, he doesn’t recognise individual people and no longer listens to one specific person. What’s more, the speech engine could still be optimised further. Amazon Echo or voice recognition systems from Apple or Google work much better, in fact.

How do people react to Pepper?

S. H.: The robot makes a fantastic impression on visitors. At the trade fair in Bregenz, our gimmicks made Pepper a real crowd-pleaser. Anyone who is interested in doing so is more than welcome to experience our robot for themselves: either at the trade fairs we attend with our Lighting Segment or in our “LightSpeed Experience Center” in Poing, Germany. Here, we demonstrate future technology related to optoelectronics. With his six lidar sensors, Pepper fits into this environment perfectly – that includes our focus on retail solutions this year.

The post Robot on-Board appeared first on Future Markets Magazine.