Automation of just-in-time delivery tasks
Automating on-demand deliveries of materials: Fetching and delivering materials from across the hospital campus to the requester
Delivery tasks in the healthcare domain present an especially well-motivated use case for robotic mobile manipulation technology.
Currently, approximately 30% of a nurse’s daily workload involves tasks that are not considered value-adding care, such as searching for people or things, food supply, maintenance of equipment and warehouse tasks (Antinaho et al., 2015).
This is of particular concern since “time nurses spend with patients is associated with improved patient outcomes, reduced errors, and patient and nurse satisfaction.” (Westbrook et al., 2011).
Autonomous mobile manipulation robots that fulfill these just-in-time delivery demands can substantially reduce the burden on facilities management personnel as well as ward nurses.
This will require robots that can search for, identify and keep track of the myriad equipment and resources in use in a hospital so as to collect and deliver them to the requester.
In addition, it necessitates generalised approaches to grasping and interacting with different, potentially novel, tools and objects, as well as adaptation for safe interaction with human co-workers and patients.
Automation of hospital bioassay sample flow
Automating patient sample flow for on-site testing or redistribution for off-site testing
Thousands of test tubes containing patient samples are processed daily by a hospital’s pre-analytical department.
This involves collecting, sorting and distributing samples to different parts of the hospital or delivering off-site for analysis. Sorting and scanning 1500 individual samples to off-site testing alone requires approximately 1.5 FTE and is typically performed by staff who are greatly overqualified for this work.
Although automated bio-assay machines can efficiently process large test batches, many of the thousands of test tubes can’t be directly put through the machines and must first be manually extracted and scanned.
This involves several repetitive steps, such as opening and closing boxes, removing and replacing lids, and sorting test tubes into trays, which is low-skilled work that also tends to cause injuries due to the repetitive physical nature of the tasks.
The Huddinge and Solna campuses of Karolinska University Hospital are the collection points for patient samples from hospitals, clinics and home testing all across Sweden. In Solna alone, 300-400 boxes arrive by courier in a single day. This equates to roughly 8000-9000 individual samples.
Currently, lab technicians must (1) immediately unpack the trays of test tubes from each arriving box so that the boxes can be returned by the awaiting couriers. (2) The trays are sorted according to (i) acute samples (urgent time-critical tests), (ii) samples that can be directly taken to the preanalytical automation without further processing, and (iii) those that lack electronic referrals and must be mannually scanned into the system, placed into compatible trays and potentially redistributed. (3) Samples from the latter set are moved to a pre-processing station before delivery to the lab.
There is great potential for robotic mobile manipulation technology to inject flexibility, reconfigurability and efficiency into the hospital bio-analysis pipeline.
Harmony aims to develop robots that can perform the sample collection, sorting and delivering tasks described above, as well as be easily programmed to interact with the various delivery containers and trays that arrive for processing.
Such a solution can provide flexible coordination with healthcare practitioners to account for critical factors such as urgency of patient results, and is also readily reconfigurable to changes in routines which naturally evolve over time.