The new system, described in the journal Human Factors, adds two haptic zones to a steering wheel at the ten o’clock and two o’clock positions. These zones can expand or contract, inflating under the driver’s palms to express the intention of the automated driving system. Crucially, the communication is not just one-directional. Drivers can respond by squeezing the haptic zones, signaling agreement or disagreement with the computer’s proposed maneuver before it happens. According to Hannah Báez, a robotics Ph.D. student and first author of the study, this ability to negotiate before an action is taken allows driver and machine to resolve conflicts in advance rather than clashing over control mid-maneuver.

To test the concept, the research team placed drivers in a driving simulation centered on an upcoming turn. The scenario was deliberately designed to create friction. Sometimes the driver and the automation would agree on which direction to turn; sometimes they would disagree. To raise the stakes further, the simulation could introduce sudden obstacles into the driver’s planned path in either situation. As the driver approached the turn, the semi-autonomous system inflated the left or right haptic portion of the steering wheel to indicate the direction it intended to go. A driver who disagreed could squeeze that portion of the wheel to register the objection and negotiate a new plan. If an obstacle appeared in the driver’s chosen path, the wheel pulsed on the corresponding side to deliver a tactile alert.

The results were striking when compared against two baselines: conditions in which drivers received no information at all, and conditions in which the automation displayed its intentions one-way, without any channel for the driver to respond. With the two-way haptic negotiation interface, driving performance improved measurably. Brent Gillespie, professor of robotics at the University of Michigan and senior author of the study, reported that drivers displayed smoother and more accurate driving paths, fought less with the automation, and used less braking while maneuvering with greater confidence. In other words, when the car and the driver could hash out a shared plan through touch, the resulting driving was simply better.

The benefits extended beyond the trajectory of the vehicle itself. Participants in the negotiation condition reported significantly lower workload, including reduced effort, frustration, and physical demand. This finding matters because mental workload is a well-documented problem in modern driver-assistance systems. When a car communicates through a barrage of auditory beeps and flashing visual alerts while simultaneously tugging at the wheel, the driver is forced to parse multiple competing channels of information under time pressure. A haptic channel embedded directly in the point of contact between driver and vehicle consolidates that communication into a single, intuitive medium, one that requires no glance away from the road and no interpretation of abstract warning symbols.

Perhaps the most consequential result concerned trust, the fragile currency on which the entire enterprise of semi-autonomous driving depends. After the automated system made a mistake, such as missing an obstacle or issuing a false warning, drivers using the negotiation interface recovered their trust in the system much faster than those relying on traditional one-way communication. The researchers attribute this resilience to the structure of the interaction itself. When driver and automation communicate intent, exchange feedback, and adjust to one another, they begin to operate as a team with transparency in decision-making, rather than as two agents taking turns behind the wheel and second-guessing each other in between.

The team is careful to note that the technology is not without risks. Some drivers became overly trusting after experiencing successful negotiations with the system, a phenomenon that raises familiar concerns about automation complacency. If a steering wheel negotiates well a dozen times in a row, a driver may come to assume it will always negotiate well, potentially disengaging their own vigilance at precisely the moments it matters most. The researchers emphasize that further study of trust dynamics is needed before such a system reaches the road, particularly around how drivers calibrate their reliance on the automation over longer exposure and across more varied driving conditions.

The technical elegance of the approach lies in its use of shared haptic channels for what human-factors researchers call shared planning and control. Conventional driver-assist systems tend to intervene silently, correcting steering without warning and leaving the driver to discover the machine’s decision through the physical sensation of resistance. That sensation of fighting for control, combined with overwhelming visual and auditory warnings, is precisely what drives some users to disable semi-autonomous features altogether. By contrast, the haptic negotiation interface makes the automation’s intentions legible before they become actions, and gives the driver a low-effort, high-bandwidth means of dissent. The wheel becomes a medium of dialogue rather than a battleground.

The implications reach well beyond the laboratory simulation. As automakers race to deploy increasingly capable driver-assistance features, the question of how humans and machines share authority has become one of the central challenges of the field. Regulatory frameworks and safety standards increasingly recognize that the interface between driver and automation is as important as the underlying algorithms. A steering wheel that can express intent, receive feedback, and support genuine negotiation offers a template for that interface, one grounded in the oldest and most trusted sensory channel a driver has: the sense of touch. The technology is covered by U.S. patent US12227190B2, applied for with the assistance of U-M Innovation Partnerships, and the team is actively seeking partners to bring it to market.

The study, titled Haptic shared planning and control: enabling coordination of future actions in human-autonomous vehicle teams through a haptic negotiation interface, was authored by Hannah Báez along with Haochi Pan, Nadine Sarter, and Brent Gillespie of the University of Michigan, and Jean Costa and John Gideon of the Toyota Research Institute. While one-way information remains the current norm for communication between autonomous vehicles and their human occupants, this research demonstrates the tangible benefits of a fuller two-way dialogue, one in which driver and automation exchange intent, agreement, and action. If the steering wheel of the future can negotiate, the drivers of the future may finally stop fighting their cars and start working with them, arriving safer, calmer, and considerably more willing to let the automation help.

Subject of Research: Haptic negotiation interfaces for communication between drivers and semi-autonomous vehicle systems

Article Title: Steering wheel communicates for better semi-autonomous driving performance

Article References: Steering wheel communicates for better semi-autonomous driving performance. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: haptic feedback, semi-autonomous driving, human-machine interaction, steering wheel interface, driver trust, automation, University of Michigan, Toyota Research Institute, Human Factors, driver-assistance systems, shared control, automotive robotics

The review followed a rigorous scoping methodology aligned with the PRISMA Extension for Scoping Reviews and the Joanna Briggs Institute Population–Concept–Context framework. The team searched three engineering and computer science databases, Scopus, the ACM Digital Library and IEEE Xplore, for publications between January 2000 and February 2025. In a striking methodological decision, the researchers deliberately excluded the word rhythm itself, along with synonyms such as timing, tempo, flow and cadence, from their search strings. Because rhythm is conceptually complex and used differently across disciplines, searching for it directly would have flooded the results with irrelevant work. Instead, they captured the broad landscape of augmentation and feedback technologies in sport, then examined whether the holistic, expert-understood concept of rhythm was actually being addressed. After screening 442 unique records from an initial pool of 529, the team distilled the literature down to 129 technology-focused studies.

The first headline finding is the era of the wearable sensor. Wearable devices were the primary technology in 48.8 percent of the included studies, and inertial measurement units, the tiny accelerometer and gyroscope packages that detect motion, appeared in 44.2 percent of the entire corpus. Their dominance reflects maturity, low cost and high ecological validity: unlike laboratory motion capture rigs, IMU-based wearables can record rhythm-relevant features such as step frequency, stance duration and inter-movement intervals directly in the field, on the track, in the pool or on the slopes. Examples in the corpus range from flexible sensor networks for sprint and jump biomechanics to smart dumbbells with embedded inertial units that evaluate strength training form in real time. For anyone hoping to quantify the temporal texture of movement, the sensor hardware problem is, in essence, already solved.

The second headline is a turn toward seeing rather than feeling the data. Computer vision systems accounted for 20.2 percent of primary technologies and extended reality systems for 22.5 percent, together representing more than 40 percent of the corpus. Visual capture technologies spanning ordinary RGB cameras, LiDAR and full motion capture appeared in one third of all hardware implementations. This suggests a field diverging into two camps: one that attaches discrete sensors to the body, and another that leverages optics and immersive environments where users quite literally inhabit their own performance data. The rise of extended reality, in particular, signals a shift beyond passive data logging toward experiential interfaces, exemplified by virtual reality training systems evaluated with national-level sitting volleyball players that produced measurable skill improvement.

Underpinning both camps is artificial intelligence. Explicit use of AI or machine learning to generate or adapt feedback appeared in 42 percent of the included studies. Rather than being the primary technology, machine learning frequently serves as the translation layer that converts raw sensor streams into meaningful, actionable cues. Examples range from wearable systems using deep learning to provide adaptive auditory feedback for weight training to computer vision pipelines that assess landing form in basketball. In principle, this is exactly the kind of computational machinery needed to move from simple real-time alerts toward the sophisticated, personalized, adaptive feedback systems that the authors argue elite sport actually requires.

Yet the feedback side of the equation reveals a striking monoculture. Visual feedback dominated the literature, appearing in 81 percent of studies, most commonly as on-screen graphs and numerical displays delivered via smartphone or tablet in 61 percent of the corpus. A representative system for hammer throw training gives coaches a real-time plot of the wire tension curve, elegant for a coach watching but demanding that an athlete look away from their own movement. The authors point out that visual attention during performance is a limited and critical resource, and studies on augmented reality workout systems have found that on-screen cues can impose additional cognitive load during dynamic movement. For an embodied skill like rhythm, which is often felt as much as seen, defaulting to screens may be actively counterproductive.

The untapped opportunity, the review argues, lies in the senses sport has largely ignored. Auditory feedback appeared in only 25 percent of studies and haptic feedback in a mere 11 percent. Their temporal affordances make them natural candidates for rhythm cueing: the ear is a timing organ, and the skin can feel a pulse without any attention being paid. The few studies that do explore these channels hint at real promise. Interactive sonification, in which movement is converted into continuous sound, has provided effective low-latency feedback on body roll in swimming, and electrical muscle stimulation has been used to directly correct running foot-strike patterns. Crucially, the auditory channel often remains unengaged during high-concentration physical exercise, leaving bandwidth free for guidance that refines rather than disrupts an athlete’s flow.

The review also documents a troubling gap between who builds these systems and who is supposed to benefit from them. Elite and professional athletes were the target population in only 26 percent of studies, and coaches in just 18.5 percent, while novices and amateurs, often university students, are heavily over-represented. The most common application domain was general fitness training, and 70 percent of studies targeted posture or technique form, compared with only 35 percent targeting timing or tempo, 20 percent targeting consistency or flow, and 12 percent targeting stride parameters. In other words, the literature measures discrete kinematic snapshots when rhythm is an integrated temporal quality. Elite performers possess refined, idiosyncratic motor programs; their challenge is micro-adjustment of an already optimized system, requiring higher-resolution feedback than novice learning studies typically provide.

The proof problem compounds the population problem. Although 81 percent of systems were empirically evaluated, the primary goal of most evaluations was system usability or feasibility, in 45 percent of studies, while only 22.5 percent measured performance improvement and 8.5 percent measured skill acquisition. Nearly one in five studies, 19.4 percent, reported no formal evaluation at all. Perhaps most tellingly, only 26.5 percent of studies explicitly measured rhythm or a direct synonym as an outcome variable. Systems for tennis swing classification, for example, celebrated machine learning model accuracy rather than any change in player skill. The result is a literature full of working prototypes with largely unevaluated training impact.

The authors distill these observations into four challenges, concerning what is sensed, who is studied, how feedback is delivered and what proof exists, and offer them as exploratory, hypothesis-generating directions rather than evidence-based prescriptions. Their near-term proposal is compelling in its simplicity: couple the already mature wearable inertial sensing platforms with continuous, low-latency, non-visual feedback such as movement sonification or complementary haptic cues, and design evaluations that separate technical validation, user experience and genuine performance outcomes as distinct levels of evidence, a trajectory they map onto the Technology Readiness Level framework from proof-of-concept to proof-of-impact. The review has limitations, including its restriction to English-language, technology-indexed literature across three engineering databases, so field-level conclusions about rhythm training await a cross-disciplinary synthesis. But the message to sports engineers is clear: the sensors are ready, the audio and touch channels are wide open, and the athletes at the top are still waiting for technology that speaks their temporal language.

Subject of Research: Human augmentation technologies and feedback modalities for training rhythm in elite sport

Article Title: Augmenting rhythm: a scoping review of technologies and feedback modalities in elite sport

Article References: Hossain, M. T., Lindeman, R. W., Ingram, M., & Lukosch, S. (2026). Augmenting rhythm: a scoping review of technologies and feedback modalities in elite sport. Sports Engineering, 29(2), Article 32. https://doi.org/10.1007/s12283-026-00562-7

Image Credits: AI Generated

DOI: 10.1007/s12283-026-00562-7

Keywords: rhythm, elite sport, wearable sensors, inertial measurement units, real-time feedback, sonification, haptic feedback, extended reality, computer vision, artificial intelligence, sports engineering, scoping review

Teleoperation is a fundamentally unnatural task. When a surgeon sits at a robotic console, the instruments they control are separated from their hands by a chain of sensors, cables, software filters and mechanical actuators. Visual feedback arrives on a screen rather than through direct sight of the tissue, depth perception is reconstructed rather than experienced, and haptic sensation — the sense of force and touch that guides so much of manual skill — is often reduced or absent entirely. The nervous system must therefore rebuild its internal model of the body, extending it outward to encompass a machine. Researchers describe this as a form of tool embodiment, and the speed at which it happens varies dramatically between individuals.

The study focused on the dynamics of this adaptation over repeated practice sessions. Rather than measuring performance only at the beginning and end of training, the researchers tracked learning continuously, capturing how error rates, movement efficiency and coordination evolved trial by trial. This fine-grained approach revealed that learning during teleoperation is not a smooth, uniform climb toward proficiency. Instead, it proceeds in distinct phases: an early period of rapid improvement as learners discover the basic mapping between their own movements and the robot’s response, followed by a slower, more effortful phase in which the fine control needed for surgical-grade precision is gradually consolidated.

A central theme in the findings is the role of sensorimotor recalibration. When people first operate a surgical robot, their movements carry the fingerprints of a lifetime of natural tool use. They grip, rotate and translate as if holding a needle driver directly. But the robot introduces distortions: scaled motion, time lags, tremor filtering and wrist rotations that do not map one-to-one onto hand orientation. The brain must detect these systematic discrepancies and adjust its motor commands accordingly. The study shows that this recalibration follows predictable dynamics, with learners initially overcompensating for the robot’s behaviour before settling into a stable, efficient control strategy that feels increasingly intuitive.

Variability between learners emerged as one of the most striking results. Some participants adapted to the robotic interface within a handful of trials, showing immediate gains in accuracy and economy of motion. Others required substantially more repetitions before their movements stabilised, and a subset appeared to plateau at intermediate levels of skill. The researchers suggest that these differences may reflect variation in how flexibly individuals update their internal models of the tool. Prior experience with video games, laparoscopic simulation or fine manual crafts may prime the nervous system for the demands of teleoperation, although the precise contribution of such background factors remains an open question for future work.

The temporal structure of practice also mattered. Learning was strongest when sessions allowed time for consolidation between bouts of practice, consistent with a large body of evidence showing that motor memories stabilise during rest and sleep. Back-to-back practice without breaks produced faster apparent progress within a single session but weaker retention across days. This has direct implications for how surgical trainees are scheduled and assessed: a curriculum that distributes practice over time, rather than cramming simulator hours into intensive blocks, is more likely to produce durable robotic surgical skill.

Feedback emerged as another decisive ingredient. Learners improved fastest when they could immediately see the consequences of their movements — when the visual display made errors obvious and corrections possible in the next attempt. This aligns with established principles of motor learning, in which the error signal between intended and actual outcome drives updates to the brain’s predictive model. In teleoperation, however, the feedback loop is inherently more complex, because the robot itself may compensate for or amplify errors before they become visible. The study highlights the importance of designing training environments in which the true state of the instruments and the task is displayed transparently, so that the nervous system receives the cleanest possible signal for learning.

The research also speaks to a broader scientific debate about what it means to acquire expertise with a machine. Classical theories of motor learning describe the formation of an internal model — a neural representation that predicts how a tool will respond to a given command. The new findings support this framework but add nuance: during teleoperation, learners appear to build not one model but a hierarchy of them, covering the robot’s kinematics, its dynamic response and the behaviour of the remote environment itself. The layering of these representations may explain why proficiency develops in stages, and why certain aspects of robotic skill, such as suturing under magnification, take considerably longer to master than basic instrument navigation.

For the surgical profession, the practical stakes are considerable. Robotic platforms are now used in millions of procedures worldwide, yet training standards vary widely and the learning curves associated with different systems are still being mapped. Understanding the dynamics of human learning during teleoperation could help design simulation curricula that target the specific bottlenecks identified here — recalibration, feedback integration and consolidation — rather than simply accumulating hours at a console. It could also inform the development of adaptive training systems that detect, in real time, where a learner is in the trajectory from novice to expert and adjust task difficulty accordingly.

Beyond surgery, the study carries implications for any field in which humans control remote machines: piloting drones, manipulating underwater vehicles, performing remote maintenance in hazardous environments or operating robotic systems in space. In all of these domains, performance depends on the same interplay between human plasticity and machine constraints that the researchers documented. The human capacity to absorb a robot into the body’s own sensorimotor machinery is remarkable, but it is not automatic. It unfolds according to identifiable rules — rules that, once understood, can be engineered into better machines, better training and ultimately safer outcomes for the patients waiting on the other side of the console. As surgical robotics continues to expand, the science of how humans learn to wield these machines may prove as important as the machines themselves.

Subject of Research: Human motor learning dynamics and skill acquisition during teleoperation of surgical robots

Article Title: Human learning dynamics during teleoperation of surgical robots

Article References: Huang, Y., Cai, Y., Li, M., Chen, Y., & Wilson, R. C. (2026). Human learning dynamics during teleoperation of surgical robots. npj Science of Learning. https://doi.org/10.1038/s41539-026-00442-6

Image Credits: AI Generated

DOI: 10.1038/s41539-026-00442-6

Keywords: surgical robots, teleoperation, motor learning, sensorimotor recalibration, npj Science of Learning, robotic surgery training, tool embodiment, simulation, learning curves, skill acquisition, haptic feedback, internal models