Robust Running and Climbing without Feet - the Role of Tarsi

On Horizontal Surfaces

Effective foot-substrate interaction remains a major challenge in the design of legged robots. To understand the contribution of foot-like appendages to the stability of high-speed running on horizontal, hard surfaces, we studied death-head cockroaches, Blaberus discoidalis. Previous studies have demonstrated the importance of tarsal claw engagement for climbing and inverted walking, but, found no change in speed on meshes when the feet or tarsi were removed. We hypothesized that tarsi function as springs to modulate leg force, dampers to dissipate leg-substrate impact energy, and/or stabilizers to mitigate the effects of external perturbations. We compared running performance before and after complete tarsi ablation under three different conditions - flat terrain (n=5, trials=88), rough terrain (n=4, trials=76) and large lateral perturbations (n=3, trials=40). The loss of tarsi on flat terrain did not alter running speed (57.7±13.1 cm/s), stride frequency (16.3±1.5 Hz), duty factor (0.43±0.07) or ground reaction force pattern. Rough terrain locomotion, with perturbations up to three times hip height, produced no difference in running speed (36.2±18.3 cm/s, P>0.1). Neither condition provided evidence for significant force modulation, damping or stabilization. Legs appear to serve these functions. However, when subjected to large lateral perturbations, animals without tarsi were less effective at regaining heading and took longer to attain steady-state body orientation (251.3±32.3 ms) post perturbation compared to those with feet (141.0±27.8 ms). These neuromechanical responses provide new insight on appendage-substrate interactions, inspiring the design of novel feet in the next generation of legged robots. 

On Inclines 

Previously, we found that cockroaches, Blaberus discoidalis, maintain performance on flat and rough terrain even with the loss of tarsi (or feet) by relying on large spines at the tibia-tarsus joint. Here, we examined whether tarsi-less cockroaches could maintain performance on inclined surfaces. We compared running performance (average speed) on constant inclines of 0, 30, 45, 60, 75 and 90° before and after complete tarsi ablation. Loss of tarsi did not affect performance (<5% change in mean running speed) at 0, 30 and 45° inclines. Tarsi-less animals decreased speed at 60° (-54%), 75° (-87%) and 90° (-88%) relative to the intact animals. To uniquely identify the cause of failure, we ran the animals on a curved incline track (radius = 50 cm) using two surfaces, rough (700 µm beaded) and smooth (Plexi-glas), and measured failure angle. Intact cockroaches failed at 89.6±2.6° and 58.6±7.7° on rough and smooth surfaces, respectively. After tarsal ablation, animals failed at 79.0±7.9° (rough) and 21.7±6.2° (smooth). Animals with only claws ablated (rest of the tarsi intact) showed failure at 84.4±7.72° (rough) and 56.8±4.1° (smooth). Tarsi-less animals showed negligible loss in running speeds on rough surfaces. Passive spines effectively compensated for loss of tarsi or claws except at steep inclines, but on smooth surfaces contribute minimally. Probing the structure of the tarsus by isolating the heterarchical functional regimes of the different tarsal specializations is important because most engineering solutions to climbing in robots rely on a single structure and not a hybrid, robust design.


Jayaram, K, Merritt C, Full R.  2012.  Robust climbing in cockroaches result from Fault Tolerant Design using Leg Spines, 5 January. Society of Integrative and Comparative Biology Annual Meeting. , Charleston

Jayaram, K, Merritt C, Cherian A, Full R.  2011.  Running without feet - the role of tarsi during high-speed, horizontal locomotion in cockroaches, 7 January. Society of Integrative and Comparative Biology Annual Meeting. , Salt Lake City