Journalist

David J. Craig is the credited writer of Boston University's 14 November 2003 account of research connecting dynamical systems, biology and biomedical engineering. His article describes James J. Collins's work on noise and sensation, including vibrating gel insoles intended to improve the detection of weak pressure signals from the feet. This is a contribution to communicating the research; the article attributes the experimental work to the researchers.^[ref]

The scientific puzzle in that account is stochastic resonance: under suitable conditions, adding noise can improve the detection of a weak signal. A simple explanatory model is a detector that responds when $x+\eta >\theta$, where $x$ is the signal, $\eta$ the noise and $\theta$ a threshold. If $x<\theta$, a detector without noise never responds; a fluctuating $\eta$ gives a nonzero probability of crossing the threshold. The same noise can also produce a response when the signal is absent, so an increase in threshold crossings alone does not establish improved discrimination. Craig's account places this counterintuitive mechanism in the concrete setting of mechanical stimulation of the sole rather than claiming that noise always helps.^[ref]

Jennifer Chu is a science writer at MIT News. MIT's staff directory identifies her as a senior writer covering subjects including mathematics, mechanical engineering and physics.^[ref] Her credited work includes the institutional account of the MIT Cheetah robot's running-jump demonstration published on 29 May 2015.^[ref]

That account explains a scientific sequence rather than just reporting that a robot jumped: onboard lidar estimates an obstacle, the controller adjusts the approach and the motors provide the forces needed for takeoff. It also distinguishes treadmill experiments from indoor-track trials, where the available planning distance differs.^[ref]

The distinction is useful when interpreting a robotics demonstration. Detecting an obstacle and finding a mechanically feasible jump are separate tasks; a successful sensor measurement does not guarantee that a particular approaching state permits the necessary takeoff. The credited contribution here is scientific reporting of the experimental system and its conditions.