Sensor intelligence

In Security visited the UK offices of Hitachi to learn how one of their most innovative products is making society safer, more efficient and crime-free through infrared sensing that can report on our every movement while still preserving anonymity and privacy.

‘Time of Flight’ (TOF) might sound like your main concern while at an airport but it’s also an acronym at the heart of Hitachi’s 3D LiDAR (TOF) motion-sensing technology. The 3D LiDAR is a point-by-point sensor that throws up to 30 frames of IR light per second at objects within a range of 0.7-10 metres.

At an immediate level, there can be a real-time reporting on the object’s dimensions to alert on unusual behaviour that may mean a person is committing a crime or is in distress. At a broader level, a sophisticated interface can extract data trends from the received light to give managers information about activity at their site.

As he demonstrated the sensors and related application programming interface (API), Hideki Hayashi, a Business Development Expert with Hitachi, told us: “Hitachi is achieving twin goals here that for our competitors would, by definition, be self-contradictory. 3D LiDAR produces unlimited data that is both aggregated and analyzed while at the same time taking us away from the Surveillance State because the subjects are literally faceless and no cameras are involved.”

Hideki continued: “One of the most common questions that anybody in the security, safety and logistics fields asks themselves is: ‘How can I make the best possible decisions with the best possible data?’ The Internet of Things (IoT) is bombarding us with data but few products are allowing the filtering of information with the intelligence of our TOF sensors and API.”

So, what are the applications? More or less unlimited but as this technology matures, Hitachi is initially concentrating on security, retail analysis and health & safety with particular focus on staff working in hazardous areas and elderly people living alone.

3D LiDAR excels in reporting on people flow in security applications. Data from sensors can alert on atypical movement or loitering that might range from shoplifting to hostile reconnaissance by organized criminals. People-counting is a particular strength and an application where IR sensing out-performs video analytics since it is not fooled by visitors entering a venue two-abreast, in tight clusters or crossing each other’s paths just at the threshold.

3D LiDAR can report and alert on a limitless range of atypical behaviour that warrants concern by management. Use of lifts (elevators) is an obvious application; the sensors can sound an alert when maximum occupancy is reached and can also alert if a person’s posture suggests they have fallen or are in distress.

Posture analysis is at the core of Hitachi’s success with the product in the field of care for the elderly who are living alone. A trip and fall is a simple occurrence to detect and the scenarios can be far more subtle, extending to creation of an alert if a person is more than usually sedentary or is not visiting the kitchen to prepare food. As with other applications, the individual’s dignity is not compromised since the technology is in no way image-based.

The ability of the product to report on complex, often sensitive patterns of behaviour without capturing an image or identifying the subject as a recognizable individual means that the technology can allay the legitimate concerns of citizen watchdog organizations who campaign about civil liberties. Against the backdrop of CCTV usage that sees the average Londoner captured on video over 300 times a day, this IR-based sensing shows the corporate responsibility of Hitachi as it develops products that report on human behaviour to create a safer society but with appropriate checks and balances.

In the workplace, 3D LiDAR is a powerful add-on for access control and lone worker protection. The practice of tailgating can be eliminated easily. The sensors can also detect and record the precise time that a worker enters an area (notably hazardous sections of a site) and ensure that he/she is sensed subsequently at points on a designated route.

So, what format does the data come in and how does it help facility managers and other stakeholders? Data from the sensors is processed by an API that is the result of Hitachi’s collaboration with two major enterprise software companies. Reports can be in any graphic format required, with heatmaps being the most popular.

The sophisticated graphic reporting with the ability to drill down through the data to observe the behaviour of sub-groups makes 3D LiDAR particularly valuable for the retail sector, notably with premium items such as consumer electronics. Retailers install merchandized shelf edges with intelligent labelling (Intelligent LabelTM) and interactive content. Multiple TOF motion sensors then report on dwell-time as well as speed and direction of hand movement.

As he took us through some graphics, Hideki Hayashi stressed how granular the data can become. “Retailers get details such as the fact that youngsters are dwelling at a display of iPads and testing the products while their parents are, say, dividing time equally between Mac and Microsoft items. It’s an eye-opener and, of course, product displays and store layout can be changed accordingly to maximize revenue.”  Retail analysis can also be more basic, and at grocery stores with few staff and a reliance on self-scanning machines, 3D LiDAR sensors in baskets can check that the number of items selected by a shopper corresponds with the number of items eventually scanned.

A common retail application is using trends in the sensor data to make intelligent suggestions to the shopper about complementary products, a strategy that is proven to increase overall spend. Sensing and subsequently analyzing consumer behaviour does not end in-store. The sensors operate to seven metres. Ten metres is already practicable and this will soon be the specified upper range as R&D staff continue to optimize the product. Dwell-time and speed of movement can therefore be measured as shoppers are walking past large-scale digital advertising within a shopping centre or on the street. Advertisers receive precise feedback on the effectiveness of an advert including dwell-time according to hour of the day and day of week.

Analyzing dwell-time is a crucial function of the technology and it has been used effectively at a recent installation in a modern art gallery within a Tokyo mega-complex. Knowing how long visitors (or sub-sets within the general visitor profile) spend looking at a painting or sculpture helps gallery curators choose items and rationalize the layout.

3D LiDAR is adept at volume-sizing for automating dimensional measurement. It can replace machine vision video cameras leading to cost savings and improvements in accuracy. A sensor mounted above the detection place can instantly measure an object.

The sensors and API have all the functionality to measure and report on common scenarios encountered in retail, leisure and industry. In the event that a niche sector has specific needs, Hitachi offers an SDK and will work with end-users or integrators to achieve unusual measurement objectives. Background subtraction requirements can vary according to environment and the SDK allows a fine-tuned approach when ambient movement or other distracting factors are non-standard.

Security installers, integrators and system designers should note that 3D LiDAR sensors have all the mounting options you would associate with a camera. There are ceiling and wall-mounted variants and the sensors come in dust or water-proof enclosures to IP66 where ingress of contaminants is a factor. Some customers are using the sensors alongside analytics from video cameras for optimum information gathering. Installers will appreciate that individual sensors connect to the hub via standard ethernet cable, and Power Over Ethernet (PoE) is an option. Retrospective ‘piggybacked’ installation on existing structured cabling is frequent practice.

The API used with 3D LiDAR allows remote operation, management and maintenance. Data is transferred to a Cloud server during the API’s analytical process. Hitachi will continue to work with software partners to make the already intuitive API even easier to use and more versatile.

Hitachi’s 3D LiDAR (TOF) shows the company’s flexible response to an ever-increasing demand for data from security officers, safety practitioners, retailers, healthcare professionals and even legislators. Hitachi is proving agile in balancing commercial imperatives to drill down to analyse specific sub-groups against requirements to use data responsibly and protect anonymity.

For further information please visit