IEEE Future Directions
Related Articles
5G is claimed by several players in the telecom biz to be a panacea for the health care biz (Qualcomm, Ericsson, Nokia). My personal opinion is that 5G will be a good tool that will surely be used in health care, but it will be unlikely to be a game changer, given that most health care services are well supported by currently available systems.
The Health Care area is driven by growing cost concern, partly consequence of an elderly population and partly to ever more costly cure. This is further stressed by the pressure of citizens that are seeing in modern health care the possibility for better cure and healthier life. Technology, in principle, can help decreasing cost, but that requires a re-engineering of the whole area. In the first instances the need for technology increases the overall cost. The possibility to better track health parameters, provided by ambient, wearable, contact and embedded sensors implies cost for acquiring, deploying and maintaining these sensors plus a higher operational cost in analyses of the data harvested. These latter can be analysed by software, thus decreasing the cost, but still this is an additional cost!
The goal to move from a reactive cure (costly) to a proactive cure (cheaper) is not actually a system wide solution, since in the end there will be need for reactive cure and the proactive cure will be an added cost. Of course, the life span and the period in which we can stay healthy will be prolonged, hence there is no doubt on the interest and drive for this evolution.
5G home connectivity according to Nokia has a break-even point for home service around 40€ per month (that includes an external antenna to capture the network radio signal and a converter to move the signal onto the home WiFi network).
Service robots in hospices, residential floors and homes is expected to grow year over year respectively by 12%,10% and 33% driven both by lack of resources (nurses) and attempt to decrease cost. Both local and network connectivity is required to support this “health care automation”. Nokia is indicating a break even point for delivering 5G connectivity in the order of 50€ per month using dedicated network slicing, required to manage robots with low latency (around 50ms for services deployed via a network cloud). The cost of robots in a hospice is calculated between 90-900€ per month depending on the robot functionality and this makes sense with respect to the cost of equivalent human labour. These are operational cost, to which one has to factor in the cost of the robot itself. In certain areas, like ambient cleaning, drug dispensing, monitoring it is close to the break-even point. More sophisticated robots that will be able to engage in emphatic relations are still on the drawing board and have to face patient’s acceptance.
Remote patient monitoring, leveraging on sensors, is being trialled in several places, like Songdo, mostly relying on fixed network connection (plus ambient radio drop usually based on WiFi). However, this will likely be extended in patient mobility for a host of chronic conditions that are not constraining the patient indoor but nevertheless benefit from a continuous monitoring. Many chronic conditions, like diabetes, high pressure, arrhythmia, can be managed through “local” (on body) intelligence (sensors plus processing and actuators) with sporadic network connections requirements. In the future, however, with the development of BCI and DBS more stringent network connectivity requirements may become important (although never essential since a local back up will always be needed).
Local communications support, with high throughput and very low latency is surely needed for surgery mediated by robots, since the low latency is essential to provide accurate tactile feedback to the surgeon manipulating the virtual scalpel. This is also a requirement for tele-surgery, with a physical limitation to 150 km to guarantee the same tactile sensitivity as being “on site” (the target latency is in the order of 25ms with a throughput of 30/50Mbps, not quite like being “there” – which would require a latency of 1ms – but still acceptable). Longer span can also be managed accepting a decrease in tactile accurate sensation. In tele-surgery, the availability of network slicing to ensure the minimal latency is a must, hence a clear case for 5G application.
5G latency is expected to be in the order of 1ms (radio link) plus 5ms (mobile edge computing) plus 10-50ms (core network) plus electromagnetic latency (1ms each 150km –to take into account response time). This has to be taken into account when defining the service architecture and it is clear that for robot real time applications, like robot surgery most communications needs to be local.
As it can be seen by this analyses 5G is going to be needed in some very specific health care applications although in most cases the present cellular infrastructure is able to support the needed services. 5G will provide an integration of the various sensors communications links, support more distributed services architecture and offer the required network resources (and capabilities) through network slicing.
- Industry in the health care domain is expecting to leverage on 5G capabilities but it is already deploying services and infrastructures using current radio systems. A graceful transition to 5G is expected. In most cases services that absolutely require 5G capabilities will not be available before the deployment of 5G hence they will be customised to take advantage from it (differently from the manufacturing industry and vehicle industry where services requiring connectivity not provided by current radio systems are being deployed using ad hoc solutions).
